The work, carried out by Simone Amirato, Fabrizio Angiulli, and Fabio Fassetti of the University of Calabria in Italy, addresses what the authors call the Anomaly Explanation problem. Their idea is deceptively simple: an explanation of an anomaly is a compact subset of features, a subspace, in which the anomalous examples deviate most sharply from the normal population. If a fraud-detection system flags a transaction, the explanation might boil down to three or four attributes, such as transaction amount, merchant category, and time of day, that together make the record look nothing like ordinary behavior. The difficulty is that finding the best such subset is a combinatorial nightmare. With d features, the number of possible subsets grows as the power set of the feature set, a search space so vast that exhaustive enumeration is hopeless even for moderately sized problems.

RLAX, short for Reinforcement Learning-based Anomaly eXplanation, recasts this search as a Markov Decision Process. Each state in the process is a candidate subspace, and each action corresponds to adding a single feature to the current selection. The agent starts from the empty set and progressively builds up an explanation, one feature at a time. What guides it is a reward signal derived from a quantity the authors call the Subspace Margin. This margin measures the difference between the anomaly scores of known outlier examples and known inlier examples within the current subspace, normalized by the square root of the subspace size. The anomaly score itself is a classic distance-based measure: the average distance from a point to its K nearest neighbors, which serves as a proxy for local density. When the margin grows after adding a feature, the agent earns a positive reward; when the margin shrinks, the reward is negative.

To learn a policy over this enormous action space, the team turned to Proximal Policy Optimization, or PPO, a policy-gradient algorithm that has become a standard tool in the reinforcement learning repertoire. The choice is deliberate. Value-based methods such as Deep Q-Networks struggle when the number of actions is very large, and in RLAX the number of actions equals the number of features, which can easily reach into the hundreds or thousands in real applications. Policy gradient methods, by contrast, handle large action spaces gracefully. The authors used the Stable-Baselines3 implementation of PPO with a custom actor-critic policy that masks inadmissible actions, keeping the default architecture of two 64-unit tanh layers for both the policy and value networks. Once training is complete, the learned policy is distilled into a ranking of features, either through the empirical mean of discounted rewards each action collected during training or directly from the policy’s action probabilities, and candidate subspaces are read off from the top of that ranking.

A key design element is the Explanation Gain, a measure that trades off separation accuracy against explanation size. Formally, the gain of a subspace is its separation quality minus a penalty proportional to the fraction of features it uses, with a tunable coefficient alpha that defaults to 0.5. This penalty is what pushes the agent toward compact explanations. The experimental analysis shows why it matters: the subspace that maximizes raw separation, measured by the Area Under the ROC Curve, may require more than 30 percent of the available features, whereas the best-gain solution often involves only a small fraction of them while preserving comparable discriminative performance. In other words, the framework resists the temptation to throw every feature at the problem and instead delivers explanations an analyst can actually read.

The journal version of the work extends an earlier conference paper with a Subspace Refinement Strategy that makes the exploration genuinely non-monotone. In addition to add actions, the agent can now remove features it previously selected or issue a stop action to terminate the episode early. This flexibility matters for two reasons. First, remove actions let the agent backtrack out of locally suboptimal regions of the search space, revising earlier selections that looked promising in isolation but hurt in combination. Second, the stop action prevents the agent from wandering through suboptimal territory merely to exhaust its step budget. The authors impose a budget on add operations to prevent loops, but leave remove operations unconstrained since they are inherently bounded by the number of prior additions. Episodes can end through the stop action, budget exhaustion, or a patience criterion that triggers after several consecutive negative rewards.

The empirical evaluation spans benchmark datasets from the ODDS repository, including arrhythmia with 274 features, mnist with 100 features, and several others, alongside two substantially larger real-world datasets: Internet Advertisements with 1,558 features and Swarm Behaviour with 2,400 features. Against competitors including V-GAN, SOD, HiCS, a genetic subspace-selection method, and random feature selection, RLAX achieved the highest Explanation Gain on every benchmark dataset, indicating the most favorable balance between discriminative power and compactness. Statistical significance against random selection was confirmed with one-sided paired t-tests adjusted using Holm’s correction. The refinement strategy also proved its worth in training dynamics: on representative datasets it produced markedly shorter episodes, saving roughly 30 percent of execution time while achieving improved AUC at the same feature subset sizes. On synthetic scalability tests, RLAX’s runtime grew only moderately, by factors of roughly 1.5 to 1.9 as the number of instances increased twentyfold, a growth rate substantially lower than that of the genetic and SOD baselines.

Perhaps the most striking evidence for the framework’s practical value comes from a qualitative case study on the cardio dataset, which contains cardiotocography measurements of fetal heart rate with well-established clinical semantics. For a representative anomaly, RLAX selected four top features: the number of prolonged decelerations per second, the histogram variance of the fetal heart rate, the percentage of time with abnormal long-term variability, and the number of fetal movements per second. Each of these maps directly onto recognized indicators of fetal distress in the obstetric literature. Prolonged decelerations are documented as strong markers of potential fetal hypoxia and acidosis, abnormal long-term variability is one of the four fundamental characteristics used to interpret a cardiotocography trace, reduced fetal movements are a classic clinical warning sign, and heart rate variance captures deviations in oscillation amplitude in either direction. An algorithm with no medical knowledge, in other words, recovered a clinically coherent explanation.

The authors are careful about the limits of their approach. Because RLAX explores the search space through single-feature modifications, strongly non-additive interactions between features, where an informative combination only pays off after several individually unrewarding additions, can be harder to discover, though the remove actions partially mitigate this. They also note that the method assumes access to a small set of labeled anomalies and inlier examples, which in practice may come from a detector’s output or from a domain expert. Future work, they write, will explore domain-informed reward shaping, alternative separation measures, and settings with heterogeneous features and weaker supervision. The source code has been released publicly, and all datasets used in the study are openly available.

What makes RLAX more than a technical curiosity is the broader shift it represents in how machine learning systems might communicate their judgments. As automated decisions increasingly touch medicine, security, and environmental monitoring, the demand is growing not just for accurate detectors but for systems that can articulate the basis of a flag in terms a human can verify. By demonstrating that a reinforcement learning agent can navigate a combinatorially explosive search space and return explanations that align with clinical knowledge, the Calabria team offers a concrete template for that kind of accountability. The anomalies that matter most are often the ones we cannot explain away, and a tool that points precisely at the handful of features responsible turns a cryptic alarm into the beginning of an investigation rather than the end of one.

Subject of Research: A reinforcement learning framework for explaining anomalies through compact feature subspaces

Article Title: RLAX: A Reinforcement Learning-Based Framework for Explaining Anomalous Data

Article References: Amirato, S., Angiulli, F., & Fassetti, F. (2026). RLAX: A Reinforcement Learning-Based Framework for Explaining Anomalous Data. Machine Learning, 115(10), Article 246. https://doi.org/10.1007/s10994-026-07182-z

Image Credits: AI Generated

DOI: 10.1007/s10994-026-07182-z

Keywords: reinforcement learning, anomaly explanation, explainable AI, feature selection, PPO, subspace margin, anomaly detection, machine learning, interpretable AI, outlier detection, cardiotocography, high-dimensional data

The challenge they tackle is a familiar one in bioinformatics. DNA microarray datasets typically measure the expression levels of thousands of genes, yet they contain samples from only a few dozen to a few hundred patients. In this high-dimensional, low-sample-size regime, most genes are redundant, noisy, or entirely irrelevant to the disease being studied, and feeding all of them into a classifier invites overfitting, inflated computational cost, and unreliable predictions. Feature selection—the process of identifying the small subset of genes that genuinely discriminate between cancerous and healthy samples—is therefore a critical preprocessing step before any classification model is trained.

Feature selection methods come in two broad families. Filter approaches rank every feature according to a mathematical score, such as its statistical association with the class label, and keep the top performers; they are fast but blind to how the chosen features will actually behave inside a classifier. Wrapper approaches, by contrast, evaluate feature subsets by directly measuring the classification accuracy they produce, which usually yields better subsets at a much higher computational price. The new study combines the two in a pipeline: five filter techniques—ANOVA, the chi-square test, Fisher score, Gini index, and information gain—first prune thousands of genes down to the top one hundred, and then a wrapper method refines that pool to the most informative subset.

The novelty lies in the wrapper. WSO is a swarm intelligence algorithm modeled on the foraging behavior of great white sharks. A population of virtual sharks, each representing a candidate solution, moves through the search space governed by equations that mimic three shark behaviors: swimming toward prey detected by wave-like sensing, following scent trails, and converging on the position of the best-performing shark near the target. Velocity updates incorporate a constriction factor to control convergence, while parameters representing the sharks’ smelling and seeing ability govern the balance between exploration of the search space and exploitation of promising regions. The authors built a binary version, BWSO, in which each shark’s position is encoded as a vector of ones and zeros, with a one indicating that the corresponding gene is selected.

The fitness function that guides the search balances two competing goals: classification accuracy and the fraction of features removed. Because diagnostic performance matters most in a medical setting, the accuracy term receives a weight of 0.9, while feature reduction carries the remaining 0.1. The best solutions are passed forward through generations, and the authors report that this solution-based transfer strategy produces faster convergence and better feature-selection performance than a naive search.

To test the framework rigorously, the researchers constructed every possible combination of five filters, three wrappers (WSO alongside the established Ant Colony Optimization and Particle Swarm Optimization), and three classifiers (logistic regression, random forest, and support vector machine), yielding forty-five candidate models. These were evaluated on four public microarray datasets covering colon tumor, central nervous system tumors, breast cancer, and lung cancer. Because no single metric can crown a winner across datasets and classifiers, the team applied a two-stage grading procedure: models were first ranked on each of six performance measures—accuracy, precision, recall, F1 score, G-mean, and Jaccard coefficient—and then re-ranked across datasets to identify the most consistently strong performers. Friedman and Wilcoxon signed-rank tests provided statistical validation of the rankings.

The results were striking. Before any wrapper was applied, filter-based models alone performed poorly, particularly on the colon and breast cancer datasets, where accuracy rarely exceeded 90 percent—a level the authors argue is insufficient for medical diagnosis. Once the wrapper stage was added, one model dominated: GR18, which pairs information gain filtering with BWSO and a support vector machine classifier. GR18 achieved the lowest average grade on three of the four datasets, reached testing accuracies of 0.9951 on the colon dataset and 0.983 on the lung cancer dataset, and consistently recorded the highest area under the ROC curve across all four datasets. It also selected the fewest features of any top-ranked model, confirming that the shark-inspired search prunes redundancy aggressively rather than simply hoarding informative-looking genes.

Comparisons against other swarm-based wrappers reinforced the case. When the same information-gain-plus-SVM pipeline was run with cuckoo search, grey wolf optimization, firefly algorithm, and artificial bee colony wrappers, WSO delivered the best results on the colon and central nervous system datasets and competitive performance on the other two, while converging smoothly within one hundred runs. The authors note that although BWSO’s running time was not always the shortest, classification accuracy—the primary objective in healthcare analytics—was consistently superior. An ablation study fixing the population size at 30 and the accuracy weight at 0.9 showed these settings reliably produced the best trade-off between accuracy and feature reduction.

What makes the work notable is less any single accuracy figure than the demonstration that a relatively young metaheuristic, previously applied to power scheduling, wind forecasting, and cloud scheduling, transfers effectively to genomics. The authors argue that WSO’s explicit balance of exploration and exploitation helps it escape the slow convergence and poor fitness landscapes that plagued earlier feature-selection heuristics. They caution that no single model suits every cancer dataset, which is precisely why their multi-criteria grading framework matters: it offers a principled way to choose among dozens of pipeline combinations rather than relying on one metric.

The researchers see several directions ahead. The same filter-wrapper-classifier architecture could be extended to multiclass medical classification problems and to other high-dimensional domains beyond medicine, and they suggest WSO could serve real-time Internet of Things systems where computational cost is at a premium, or anchor ensemble frameworks that combine multiple feature-selection and classification models. For now, the study offers a vivid example of how borrowing the sensory strategy of an ocean predator can sharpen the search for a molecular needle in a genomic haystack.

Subject of Research: Application of the white shark optimization algorithm to feature selection and classification of high-dimensional DNA microarray cancer data

Article Title: Unveiling the potential of white shark optimization algorithm for high dimensional cancer data classification

Article References: Dash, B., & Dash, R. (2026). Unveiling the potential of white shark optimization algorithm for high dimensional cancer data classification. Discover Artificial Intelligence, 6(1), Article 1386. https://doi.org/10.1007/s44163-026-01808-w

Image Credits: AI Generated

DOI: 10.1007/s44163-026-01808-w

Keywords: white shark optimization, swarm intelligence, feature selection, DNA microarray, cancer classification, bioinformatics, machine learning, wrapper methods, filter methods, high-dimensional data, support vector machine, metaheuristics

ProD, short for “prodding” the Probability Densities, was developed by Jin Cheng Liaw, Francisco Geu Flores, and Wojciech Kowalczyk and published in the journal Machine Learning with Applications. It is a univariate filter method for feature selection, meaning it ranks features based purely on the statistical properties of the training data rather than by training and testing a specific machine learning model. That model-agnostic design carries real advantages: the ranked features can be confidently paired with any downstream algorithm, and because no iterative modeling is required, filter methods are dramatically cheaper to run than wrapper or embedded alternatives.

The core idea is borrowed from an unexpected place: ecology. Ecologists have long used the niche overlap coefficient to measure how much two coexisting species share the same resources. ProD adapts this concept to classification by estimating, for each feature, a probability density function separately for every class, and then computing the geometric intersection area between those density curves. If the distributions of two classes barely overlap on a given feature, that feature is a strong discriminator; if the curves lie almost entirely on top of one another, the feature carries little class-relevant information. The resulting score is a literal proportion between 0 and 1, which is precisely what makes the method so visually intuitive.

That visualizability is ProD’s defining departure from classical divergence metrics such as Kullback-Leibler divergence, the Bhattacharyya distance, and the Hellinger distance. These traditional measures remain popular largely because of mathematical convenience, offering closed-form solutions under Gaussian assumptions and, in the case of KL divergence, smooth differentiability suited to gradient-based optimization. But they are abstract, hard to plot, and prone to numerical instabilities when estimated non-parametrically. KL divergence, for instance, can blow up through division by zero in the tails of kernel density estimates. ProD sidesteps these problems by combining kernel density estimation with continuous numerical integration, producing a bounded, directly interpretable geometric quantity that faithfully reflects complex, multi-modal distributions without forcing parametric assumptions.

The method also distinguishes itself from mutual information and Relief-based filters, two other pillars of the feature selection literature. Mutual information, grounded in Shannon entropy, measures dependency between feature and target in bits or nats; Relief-family algorithms assign weights by repeatedly comparing each instance’s feature values with those of nearest neighbors of the same and different classes, effectively computing a spatial margin. Both are well established, but their outputs are proxy statistics that domain experts often find difficult to interpret. Because ProD evaluates the physical overlap area of continuous densities, practitioners can validate a feature ranking at a glance by simply plotting the class distributions, rather than trusting an opaque number.

Technically, the pipeline is straightforward. Each feature is min-max normalized to the interval from 0 to 1, split into class-specific subsets, and smoothed with Gaussian kernel density estimates along a fine evaluation grid. Bandwidth selection, which controls how much the densities are smoothed, is automated using Scott’s or Silverman’s classic rule-of-thumb formulas, with Scott’s rule emerging as the better default in testing. The intersection area between every pair of class densities is then computed by numerical integration and averaged across all class pairs. The authors also built in safeguards for awkward data: when the interquartile range collapses to zero on highly discrete or skewed arrays, the algorithm falls back on the sample standard deviation, and grid tails can be extended to handle boundary-heavy data.

Benchmarks were extensive. On synthetic datasets engineered with known relevant, redundant, and irrelevant features, including circuit-inspired ANDOR and ADDER problems and simulated microarray data with 4,060 features, ProD using Scott’s rule achieved the highest average index of success at 76.90, edging out established competitors such as ReliefF, I-Relief, LH-Relief, mutual information, the ANOVA F-ratio, and mRMR, while remaining highly robust across repeated dataset iterations. On real datasets spanning cancer microarrays, epilepsy electroencephalograms, gait recognition, human activity recognition, and network intrusion detection, ProD performed comparably to the strongest methods while running orders of magnitude faster than Relief-based approaches, which scale quadratically with sample size. ProD’s theoretical complexity grows only linearly in both samples and features, and its per-feature evaluation parallelizes cleanly across processor cores.

The method is not without limitations. As a univariate filter, it evaluates each feature in isolation and does not deliberately remove redundancy or capture multi-feature interactions, a deliberate trade-off that buys complete immunity to the curse of dimensionality and total ranking stability even when massive numbers of noisy features are added. Its reliance on kernel density estimation also makes it sensitive to sample size, with the authors warning users when a class contains fewer than five observations. On purely discrete data, probability mass functions or information-theoretic metrics remain theoretically more appropriate, and experiments confirmed that ProD’s selection ability is less reliable in discrete feature spaces unless bandwidths are carefully tuned.

One of the most instructive findings concerns multiclass problems. ProD’s default setting averages overlap areas over all pairs of classes, which implicitly assumes that an informative feature must separate every class pair simultaneously. On the synthetic ADDER dataset and the human activity recognition dataset, this default penalized features that brilliantly distinguished broad activity categories but failed on specific confusing sub-pairs, such as sitting versus standing. Switching the combinatorial hyperparameter from pairwise comparisons to evaluating all classes at once produced dramatic improvements, including perfect success rates on the discrete ADDER datasets and top-tier accuracies on the activity recognition task. A built-in heatmap of pairwise overlap areas lets users diagnose exactly which class pair is causing a feature to fail.

Beyond raw performance, the authors position ProD as a tool for human-in-the-loop machine learning, an increasingly urgent need in high-stakes domains like healthcare and drug discovery where black-box models have struggled to deliver on their promise. Because the method renders the underlying class distributions of every top-ranked feature, users can spot outliers that may be mislabeled samples, or peculiar clusters hinting at systematic errors in data collection, and iteratively refine both the dataset and the model. In interactive machine learning, where fast iteration often matters more than optimal first-pass results, a filter that is simultaneously fast, model-agnostic, and fully visualizable may prove to be exactly the kind of transparent interface that lets domain experts, not just algorithms, steer the science.

Subject of Research: A visualizable filter feature selection method based on class probability density overlap

Article Title: ProD: A visualizable filter-feature selection method based on “prodding” the class {Pro}bability {D}ensities for overlappping

Article References: ProD: A visualizable filter-feature selection method based on “prodding” the class {Pro}bability {D}ensities for overlappping. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: feature selection, machine learning, kernel density estimation, class overlap, niche overlap coefficient, filter methods, explainable AI, human-in-the-loop, high-dimensional data, bioinformatics, statistics, open-source software

The Southeastern Universities Research Association is a consortium of research universities that works to advance collaborative research and education across the region and to strengthen the scientific enterprise of the United States more broadly. The organization supports research operations and workforce development in partnership with member universities, government agencies and national laboratories. Its Graduate Awards program is designed to identify and support emerging STEM talent at a pivotal moment, when students are transitioning from master’s-level training into the sustained, independent inquiry that defines doctoral research. Each recipient receives 5,000 dollars to support professional development and to help reduce the financial barriers that often accompany the start of doctoral study, a stage at which promising researchers can otherwise be forced to divide their attention between scholarship and outside employment.

Anab’s research trajectory illustrates the increasingly practical orientation of modern computer science graduate training. During his master’s studies at Georgia State, he applied data science and artificial intelligence techniques to real-world datasets, developing webcam-based systems capable of stress signal detection, emotion recognition and real-time sign language translation. These projects sit at the intersection of computer vision, machine learning and human-computer interaction, requiring models that can interpret subtle visual cues from video streams in real time rather than analyzing static images after the fact. Webcam-based sensing is an area of growing interest in digital health because it turns an ordinary camera already present in most laptops and phones into a passive monitoring instrument, avoiding the cost and friction of dedicated wearable sensors while raising important questions about accuracy, robustness and privacy that researchers in this field must confront directly.

As a doctoral researcher, Anab plans to build on that foundation by developing AI-enabled digital health technologies. The ambition is to move from laboratory demonstrations toward practical tools that can improve how human well-being is monitored and understood, combining advances in artificial intelligence, sensing and digital health into systems that are reliable enough for everyday use. In his own words, receiving the SURA Graduate Award is a meaningful recognition of his research journey and a motivation to continue developing technologies that can have a real impact on human health, with support that will help him advance doctoral research aimed at creating practical tools for monitoring well-being. The transition from detecting stress signals in controlled settings to deploying dependable health technologies is a demanding one, and sustained fellowship support of the kind the SURA award provides is often what allows doctoral students to pursue it rigorously.

Cui’s work addresses a different but equally consequential frontier: the statistical theory that underpins machine learning and data science. During his master’s research in Georgia State’s Department of Mathematics and Statistics, he developed new multivariate modeling approaches for high-dimensional datasets, work that has produced a first-author paper currently under review. High-dimensional data, in which the number of variables approaches or exceeds the number of observations, is now the norm in fields ranging from genomics to finance to large-scale behavioral studies, and it poses a fundamental challenge to classical statistical methods that were designed for settings with many more data points than parameters. Cui has described what excites him most as building statistical methods that stay reliable when data are high-dimensional and messy, which is precisely the regime in which machine learning tools are hardest to trust. His selection as one of two students chosen for the department’s machine learning concentration reflects the growing recognition within mathematics departments that rigorous statistical theory is essential to the credibility of modern artificial intelligence.

The distinction between Cui’s theoretical orientation and Anab’s applied systems work captures two complementary halves of the contemporary data science enterprise. Applied AI systems such as webcam-based health monitors depend on statistical guarantees to behave predictably when confronted with noisy, shifting, real-world inputs, while theoretical advances in multivariate modeling gain relevance when they inform the design of learning algorithms that must operate on imperfect data at scale. That both researchers emerged from the same college at the same university, and were recognized in the same year by the same consortium, suggests how deeply data-driven methods now permeate disciplines that were once sharply separated, from computer science departments focused on engineering working systems to statistics departments focused on inference and uncertainty.

Both students are based in Georgia State’s College of Arts & Sciences, and their recognition carries institutional weight beyond the individual honor. Donald Hamelberg, the university’s Vice President for Research and Economic Development and himself a Georgia State doctoral alumnus, attended the ceremony honoring this year’s awardees at the SURA Annual Members Meeting and accepted Cui’s award on his behalf. Having two Georgia State students among the four awardees, he noted, speaks to the strength of the university’s graduate research community, and he expressed pride in seeing the two students’ efforts recognized as a reflection of the talented researchers advancing discovery at the institution. For a university that has invested heavily in building research capacity across the computational and quantitative sciences, the concentration of awards in artificial intelligence and statistics is a visible signal of where that investment is bearing fruit.

The selection process itself was led by M. Omar Faison, chair of the SURA Committee for Collaborative Capacity Building, who oversaw the evaluation of candidates from member universities. He described the students as selected for their academic excellence and dedication to advancing research as outstanding members of the next generation of STEM talent, and emphasized that the SURA Graduate Award offers awardees an added measure of support during a key moment in their research journeys. The framing matters: early-career awards of this kind do more than provide money. They signal to doctoral advisors, hiring committees and funding agencies that a researcher has been independently vetted by a multi-institutional body, and they give recipients the freedom to take intellectual risks in the first years of a doctorate, when the direction of a research career is often set.

The setting of the recognition added its own symbolic resonance. Anab and Cui were honored at the 2026 SURA Annual Members Meeting held at the Thomas Jefferson National Accelerator Facility, a U.S. Department of Energy national laboratory in Newport News, Virginia, known widely as Jefferson Lab. The facility, which probes the fundamental structure of matter using high-energy electron beams, represents the kind of large-scale, collaborative scientific infrastructure that consortia like SURA exist to connect with university research programs. Honoring emerging researchers in artificial intelligence and statistics at a national laboratory dedicated to experimental physics underscores how the boundaries of STEM research continue to blur, with computational and statistical expertise now indispensable to nearly every scientific domain, from particle physics to digital health.

For the two awardees, the award arrives as they begin doctoral work that will unfold over the coming years, in fields that are evolving almost faster than curricula can be updated. Anab’s plan to develop AI-enabled digital health technologies places him in one of the most competitive and fastest-growing areas of computer science, where the difference between a compelling prototype and a deployable clinical tool often comes down to years of careful validation. Cui’s focus on statistical theory for high-dimensional machine learning addresses a foundational vulnerability of the AI systems now reshaping science and society, since methods that cannot be trusted on messy data cannot be trusted at all. The SURA recognition, and the financial support that accompanies it, gives both researchers a measure of runway at exactly the stage when sustained focus matters most, and it offers an early, externally validated vote of confidence in the directions they have chosen to pursue.

Subject of Research: SURA 2026 Graduate Awards recognizing two Georgia State doctoral students in artificial intelligence and statistical modeling

Article Title: Two Georgia State students recognized among emerging STEM leaders

Article References: Two Georgia State students recognized among emerging STEM leaders. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: SURA, Georgia State University, graduate awards, STEM, artificial intelligence, digital health, computer vision, statistics, machine learning, high-dimensional data, doctoral research, Jefferson Lab

The central theme of this year’s meeting is the mathematics of data science in high dimensions, a phrase that sounds technical but describes one of the most consequential challenges in contemporary computation. Real-world datasets, whether they describe gene expression profiles, pixel intensities in images, or the embeddings that large language models use to represent words, live in spaces with thousands or millions of coordinates. Human intuition, forged in a world of three dimensions, fails spectacularly in such settings. Distances behave strangely, volumes concentrate in unexpected places, and phenomena that seem impossible at small scale become routine. The mathematics needed to navigate these spaces, from concentration inequalities to random matrix theory, has become the invisible scaffolding of modern artificial intelligence.

Dimensionality reduction and embeddings form one of the announced focal topics, and for good reason. Techniques such as principal component analysis, random projections, and manifold learning all attempt the same fundamental task: compressing enormous datasets into smaller representations that preserve the structure that matters. The theoretical guarantees behind these methods, including the Johnson-Lindenstrauss lemma and its many descendants, tell practitioners exactly how much information can be retained and at what computational cost. As embedding-based systems have become the backbone of search, retrieval, and generative models, the demand for sharper theory about when these compressed representations are faithful, and when they silently distort, has grown correspondingly urgent.

Scalability and parallel algorithms constitute a second announced focus, reflecting a reality that every practicing data scientist knows intimately: a beautiful algorithm that cannot run at scale is, for most purposes, not an algorithm at all. The conference program will address how classical numerical methods must be rethought for distributed and parallel architectures, where communication costs between processors often dominate the raw arithmetic. Questions of numerical stability, synchronization, and the trade-offs between exact and approximate computation take on new life when datasets no longer fit on a single machine. Work in this area connects the oldest traditions of numerical analysis with the newest hardware, from GPU clusters to specialized accelerators.

Algebraic and geometric data analysis rounds out the highlighted topics, and it is here that the meeting’s intellectual ambitions are perhaps most visible. Topological data analysis, tensor decomposition, and algebraic approaches to statistical problems treat data not as a table of numbers but as an object with shape and structure. A cloud of points may reveal a loop, a void, or a manifold whose topology explains a hidden variable in the data-generating process. Tensors, the higher-dimensional generalizations of matrices, have become essential for representing multi-way data in neuroscience, chemometrics, and machine learning, and their decomposition theory draws on deep results in multilinear algebra. These methods remain mathematically demanding, which is precisely why a conference devoted to their foundations matters.

MDS26 will not stand alone. The meeting will be held jointly with the SIAM Conference on Imaging Science, IS26, and the SIAM International Conference on Data Mining, SDM26, creating a triple convergence of communities that rarely have the chance to interact at this scale. Imaging science brings its own mathematical culture, built on inverse problems, variational methods, and the physics of measurement, while data mining contributes a strongly applied tradition of pattern discovery, clustering, and graph analytics. Holding the three conferences together means that a single registration opens the door to sessions spanning the full pipeline from sensor and image formation through mathematical modeling to deployed analytical systems.

The joint format reflects a deliberate philosophy about how progress in data science actually happens. Breakthroughs rarely emerge from a single discipline working in isolation. Compressed sensing, one of the defining results of the last two decades, required harmonic analysis, optimization, and signal processing to advance together. The theory of neural networks is being reshaped today by statisticians, dynamical systems experts, and approximation theorists working in parallel. By colocating imaging, data mining, and the mathematics of data science, the organizers have engineered the kind of cross-pollination that historically produces such breakthroughs, allowing a statistician wrestling with high-dimensional inference to wander into a session on inverse problems in tomography and find the missing piece.

Presentations at the meeting will span the full range from foundational theory to real-world applications, according to the announcement. That spectrum is a defining feature of the SIAM conference series, which has long served as a bridge between communities that publish in different journals and attend different meetings. Theoretical results on optimization landscapes or the convergence rates of sampling algorithms may appear on the same program as case studies from national laboratories or industrial research groups. The stated aim is to highlight advances in mathematical, statistical, and computational methods that shape how data are analyzed, modeled, and used to inform decision-making, a formulation that deliberately places decision-making, not computation, at the endpoint of the pipeline.

For attendees, the practical significance of the meeting lies in its timing. The past several years have seen artificial intelligence systems deployed at a pace that has outstripped the theory meant to explain them. Understanding why large models generalize, when embeddings can be trusted, and how to certify the outputs of high-dimensional statistical procedures are open problems with immediate consequences. Conferences like MDS26 function as the venues where the mathematical community organizes its response, identifying which empirical observations deserve theoretical explanation and which theoretical tools are ready to be transferred into practice. The organizers extend an explicit invitation for participants to engage with emerging ideas, share insights, and help define the next generation of mathematics for data science.

Registration for MDS26 is open now through the Society for Industrial and Applied Mathematics, with full details for the conference and its two joint meetings, IS26 and SDM26, available on the SIAM conference website. For researchers whose work lives in the high-dimensional spaces where modern data resides, and for practitioners who depend on the guarantees that mathematics provides, the meeting offers a concentrated survey of where the field stands and where it is heading. In a discipline that increasingly determines how societies analyze, model, and act on information, the mathematics of data science has never mattered more, and MDS26 is positioned as one of the year’s essential gatherings for anyone building its foundations.

Subject of Research: Mathematical foundations of data science, including high-dimensional data analysis, dimensionality reduction, scalable algorithms, and algebraic and geometric methods

Article Title: SIAM Conference on Mathematics of Data Science (MDS26)

Article References: SIAM Conference on Mathematics of Data Science (MDS26). (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: SIAM, MDS26, mathematics of data science, high-dimensional data, dimensionality reduction, embeddings, parallel algorithms, algebraic data analysis, geometric data analysis, imaging science, data mining, conference

The core difficulty the researchers set out to solve lies in a family of techniques known as Multi-Attribute Decision-Making, or MADM. These methods are designed to rank alternatives—here, scenic spots—against a set of weighted criteria. They work elegantly when the number of attributes is small and the data are homogeneous. But modern tourism evaluation data, harvested at scale from platforms such as Ctrip, violate both assumptions. Hundreds of attributes may be involved, many of them correlated, some redundant, and some outright contradictory in how they rank the same destinations. Feeding such high-dimensional data directly into a decision model can distort results, because redundant or conflicting attributes exert an outsized influence on the final score.

The team’s solution unfolds in three stages. The first stage is a consistency-driven dynamic clustering algorithm. Instead of clustering the scenic spots themselves, the method clusters the evaluation attributes, grouping together those criteria that produce similar ranking patterns across the destinations being compared. In other words, if two attributes—say, visitor ratings of cleanliness and ratings of facility maintenance—consistently order the scenic spots in nearly the same way, they are placed in the same cluster. The word dynamic is important: the algorithm adjusts cluster assignments as it processes the data rather than fixing them in advance, and it uses measures of rank consistency to decide when attributes genuinely belong together. This step reduces the dimensionality of the problem in a principled way, without discarding any information outright.

The second stage applies a well-established decision tool within each cluster: TOPSIS, the Technique for Order Preference by Similarity to Ideal Solution. TOPSIS, introduced in the operations research literature in the 1980s and refined through the 1990s, ranks alternatives by measuring how close each one is to an ideal positive solution and how far it is from a negative ideal solution. Within a single cluster of similar attributes, this is a reasonable and well-conditioned calculation, because the attributes behave coherently. Running TOPSIS separately inside each cluster yields a set of local rankings—one ordered list of scenic spots per cluster—each reflecting the perspective of a coherent family of evaluation criteria.

The third stage is where the method borrows from an unexpected field: voting theory. The researchers use the Borda count, an aggregation procedure devised in the eighteenth century by Jean-Charles de Borda, to fuse the local rankings into a single global ranking. Under the Borda count, each local ranking awards points to each scenic spot according to its position, and the spot with the greatest accumulated total across all clusters wins the top place. This is the same arithmetic that determines winners in some ranked-ballot elections, and its appeal here is the same as in politics: it aggregates many partial, imperfect orderings into one socially robust verdict while dampening the influence of any single anomalous cluster. The authors note that prior theoretical work has argued the Borda count is in an important sense the optimal ranking method, which makes it a natural choice for fusion.

The pipeline was tested on evaluation data for China’s 5A scenic spots, collected from publicly accessible tourism review information on the Ctrip platform. The 5A designation is the pinnacle of China’s tourist rating system, so the destinations involved are already elite; the question was whether a data-driven ranking built from raw, high-dimensional visitor evaluations could reproduce and refine the official assessments. The empirical results showed a high degree of consistency between the method’s output rankings and existing tourism ratings, suggesting that the clustering-and-fusion pipeline captures the same underlying quality signals that expert evaluators respond to, but derives them transparently from visitor data rather than from a fixed administrative rubric.

Why not simply cluster the destinations instead of the attributes? The authors’ design choice reflects the structure of the problem. Tourists evaluate the same scenic spot along many dimensions, and the statistical relationships among those dimensions—how they correlate, which ones move together across destinations—carry the real information. By grouping attributes with similar ranking behaviour, the method effectively discovers latent evaluation perspectives: bundles of criteria that visitors treat as a single experiential quality. Each cluster then functions as a coherent judge, and the Borda count lets these judges vote. This architecture mirrors recent work on large-scale group decision-making, in which clustering and consensus techniques are used to manage hundreds or thousands of participants, and it connects to a growing literature on ranking fusion for massive online review datasets.

The technical payoff is robustness against heterogeneity. Real-world evaluation data are rarely generated by a single consistent process: different visitors weight different concerns, different attributes suffer different amounts of noise, and some criteria may be systematically inflated or deflated on particular platforms. A monolithic MADM model must average all of these effects together, which can wash out genuine distinctions between destinations. The cluster-then-rank-then-fuse design isolates each coherent evaluation channel, ranks destinations within that channel, and only then combines the verdicts. Conflicting attributes are no longer forced into a single weighted sum; they are allowed to speak in their own cluster and then reconciled at the voting stage. The authors position the framework as particularly valuable for scenarios involving high-dimensional and heterogeneous evaluation data, where conventional approaches struggle.

The practical implications extend beyond holiday planning. Tourism management authorities could use such a framework to monitor how the data-driven ranking of attractions evolves as visitor sentiment shifts, identifying destinations whose reputations are improving or deteriorating and tracing those changes back to specific clusters of attributes. Travel platforms could integrate the method into recommendation engines that must reconcile millions of reviews across many criteria. And the underlying algorithmic idea—cluster the criteria, rank within clusters, fuse by voting—is domain-agnostic, potentially transferable to any multi-attribute decision problem drowning in high-dimensional data, from product selection to service quality assessment.

The study also offers a candid account of its data foundations. The datasets were drawn from publicly accessible tourism evaluation information on Ctrip, and the full raw dataset cannot be redistributed because of third-party platform terms of use and intellectual-property restrictions. To support reproducibility, the published paper documents the data source, structure, preprocessing, variable construction, clustering procedure, parameter settings and experimental design in detail, with derived statistical information and non-restricted processed data available from the corresponding author on reasonable request. The research was supported by the Hunan Provincial Natural Science Foundation of China and the Scientific Research Fund of the Hunan Provincial Education Department. As online reviews continue to swell into one of the largest bodies of human judgement ever collected, methods like this one—part statistics, part social choice theory—point toward a future in which the wisdom of crowds is not merely accumulated, but properly counted.

Subject of Research: A dynamic clustering and ranking fusion method for multi-attribute evaluation of 5A scenic spots using tourist review data

Article Title: Dynamic clustering and ranking fusion for tourism recommendations: A data-driven approach to 5A scenic spot evaluation

Article References: Li, W., Xiao, Z., Yu, B., & Xu, Z. (2026). Dynamic clustering and ranking fusion for tourism recommendations: A data-driven approach to 5A scenic spot evaluation. Applied Intelligence, 56(15), Article 433. https://doi.org/10.1007/s10489-026-07488-4

Image Credits: AI Generated

DOI: 10.1007/s10489-026-07488-4

Keywords: tourism recommendation, dynamic clustering, ranking fusion, TOPSIS, Borda count, multi-attribute decision-making, high-dimensional data, 5A scenic spots, online reviews, Ctrip, consistency-driven clustering, Applied Intelligence

The method, called FS-GSR—Feature Selection with Group Streaming via Graph-based Scoring and Sparse Reconstruction—was developed by Tianyuan Jia of The Hong Kong Polytechnic University. It tackles a core weakness of existing streaming feature selection techniques: most process features one at a time, ignoring the structural relationships and redundancies that exist both within and between groups of features. When features arrive in batches, as they do in multi-view learning systems or modular sensor deployments, sequential individual processing can disrupt group structures and degrade performance. FS-GSR instead operates at the group level, updating the selected feature subset in real time as each new batch arrives.

At the heart of the approach is a three-stage framework that functions like a computational funnel. The first stage performs a hybrid, graph-guided filtering of each incoming feature group. The algorithm constructs a similarity graph over the samples, connecting points that are close in the feature space, and computes a Laplacian Score for every feature. This score measures how smoothly a feature varies across locally similar samples—features that preserve the intrinsic geometric structure of the data receive low scores, while noisy, irregular features are penalized. In parallel, a supervised F-score evaluates each feature’s discriminative power by comparing between-class variance to within-class variance. The two measures are blended with equal weight, and only the top fraction of features by combined score survives this initial screening.

The second stage refines this candidate pool by preserving structure at a finer granularity. The algorithm performs a spectral decomposition of the graph Laplacian, extracting ten eigenvectors that form a low-dimensional embedding of the sample manifold. Each surviving feature is then evaluated by how well it can reconstruct this structural representation, using a closed-form projection solution that avoids costly iterative optimization. Features that best approximate the manifold’s skeleton are retained, ensuring that the selected subset captures not just individual relevance but the collective geometry of the data. Because the reconstruction coefficients have an explicit analytical solution, this stage adds minimal computational overhead.

The third and final stage is where FS-GSR departs most sharply from its predecessors. All candidate features accumulated from the first two stages are merged into a global pool, and a multi-task Lasso regression is applied across the entire set simultaneously, using one-hot encoded class labels as the response. The L2,1-norm penalty drives the coefficients of redundant features to exactly zero, yielding a compact, non-overlapping final feature set. The authors also provide theoretical backing for this stage, formulating it as a row-support recovery problem in high-dimensional multivariate regression. Under standard conditions on the covariance structure, they prove that the sparse reconstruction can recover the true set of informative features with high probability, provided the effective sample complexity exceeds a specific threshold.

The computational design is notable for its scalability. By confining expensive cubic operations to the fixed sample space rather than the growing feature dimension, the algorithm’s cost per streaming group scales linearly with the number of incoming features. Because the first two stages aggressively compress each group before global optimization, the feature pool fed into the multi-task Lasso remains small, preventing the exponential blow-up that plagues conventional structure-preserving methods. Scalability experiments on the Ovarian Cancer dataset confirmed near-linear growth in running time as feature dimension increased, and revealed a V-shaped trade-off in execution time as the number of streaming groups varied—very few large groups incur heavy intra-group matrix operations, while very many small groups accumulate loop overhead.

Empirically, the method was tested on five benchmark datasets spanning biomedical, image, and text domains, including Colon (62 samples, 2,000 features), Ovarian Cancer (253 samples, 15,154 features), MLL, COIL20, and PCMAC. These datasets deliberately covered contrasting correlation structures: the Ovarian Cancer data exhibited strong intra-group correlations of up to 0.88, while the PCMAC text dataset showed extremely sparse correlations of roughly 0.025. Against four established baselines—OGFS, Group-SAOLA, Fast-OSFS, and alpha-investing—FS-GSR consistently achieved the highest or near-highest classification accuracy and F1 scores under both KNN and SVM classifiers. On the Ovarian Cancer dataset, it reached accuracy and F1 values above 0.99.

Perhaps the most striking result concerned feature compression. In one experimental configuration, FS-GSR retained an average of just 8.12 features from the 15,154-dimensional Ovarian Cancer dataset, while simultaneously improving KNN accuracy from 0.9845 to 0.9960 and SVM accuracy from 0.9885 to 0.9980—a 63 percent compression relative to the two-stage variant alone. Ablation studies confirmed that each stage contributes distinct benefits: the first enables efficient local filtering, the second improves structural consistency, and the third enforces global sparsity and discriminative refinement. The authors note that the degree of compression adapts to each dataset’s intrinsic correlation structure, with strongly correlated data requiring more retained features to preserve associative signals.

Parameter sensitivity analyses reinforced the method’s practicality. Classification accuracy remained stable across broad plateaus for the retention ratios governing the first two stages, and the mixing weight balancing structural and discriminative scoring performed robustly between 0.2 and 0.8, peaking near 0.5. The embedding dimension saturated quickly, with accuracy plateauing once it exceeded five, justifying the default setting of ten. This robustness means practitioners can deploy the method without exhaustive grid searches, using a simple boundary-driven tuning strategy on validation data.

The implications extend well beyond the benchmark datasets. Streaming, group-wise feature arrival is characteristic of stepwise sensor deployment, progressive module activation in multi-view systems, and sequential extraction pipelines in bioinformatics and text mining. By combining graph-based manifold learning with sparse global reconstruction in an online framework, FS-GSR offers a template for learning systems that must remain both accurate and parsimonious as data evolves. The author identifies future extensions toward multi-label settings, incomplete data systems, and broader generalization, suggesting that the challenge of features that never stop flowing is one the field is only beginning to master.

Subject of Research: Group streaming feature selection for high-dimensional data using graph-based scoring and sparse reconstruction

Article Title: FS-GSR: Graph-based scoring and sparse reconstruction for group streaming feature selection

Article References: Jia, T. (2026). FS-GSR: Graph-based scoring and sparse reconstruction for group streaming feature selection. Applied Intelligence, 56(15), Article 445. https://doi.org/10.1007/s10489-026-07498-2

Image Credits: AI Generated

DOI: 10.1007/s10489-026-07498-2

Keywords: feature selection, streaming features, graph-based scoring, sparse reconstruction, multi-task Lasso, Laplacian score, machine learning, high-dimensional data, data mining, classification accuracy, dimensionality reduction, Applied Intelligence

The central obstacle the researchers set out to overcome is a computational one that has shaped the field for over a decade. Classical subspace clustering methods rely on a self-representation model, in which every data point is expressed as a linear combination of all the other points. The resulting coefficient matrix encodes the relationships that define cluster membership, and it is extraordinarily informative. The problem is its size. For n data points, the full sample-to-sample affinity matrix contains n squared entries, so a dataset of one million samples would demand a matrix with a trillion coefficients. Constructing, storing, and optimizing over such an object is prohibitive on any hardware, which is precisely why many theoretically elegant subspace clustering algorithms have remained confined to benchmark datasets of a few thousand points.

The LSSC framework breaks this bottleneck by replacing the full self-representation with a compact sample-to-anchor representation. Instead of allowing every point to be reconstructed from every other point, the method first selects a much smaller set of anchor points that serve as representative landmarks for the data. Each sample is then expressed as a nonnegative combination of these anchors alone. Because the number of anchors grows far more slowly than the number of samples, the coefficient matrix shrinks from n squared to a size proportional to n times the anchor count, turning an intractable optimization into a manageable one. Crucially, the nonnegativity constraint on the coefficients gives the representation a natural probabilistic flavor: each weight reflects how strongly a sample belongs to the neighborhood of a given anchor, which makes the downstream clustering step far more stable.

Compactness alone, however, is not enough. The authors identify a second, subtler failure mode in existing scalable methods: fast representations often fail to preserve the intrinsic structural relationships of the data when distributions become complex. Two points that sit close together on the same underlying manifold may end up with very different anchor coefficients simply because of how the anchors were sampled, and the resulting affinity graph can fragment genuine clusters. LSSC counters this with a locality-aware coefficient initialization, which seeds the optimization using neighborhood information so that the reconstruction process starts from a configuration already consistent with the local geometry of the data. This initialization acts like a well-chosen starting point in a rugged optimization landscape, steering the solution toward coefficients that respect the manifold rather than artifacts of sampling.

On top of that initialization, the framework applies distance-weighted structure regularization, a mechanism that explicitly encourages the learned representation to remain smooth with respect to the geometry of the data. Samples that are near each other in the original feature space are pushed toward similar anchor coefficients, weaving locality information directly into the optimization objective rather than bolting it on afterward. The framework also employs coefficient regularization to keep the representation compact and well-conditioned, suppressing degenerate solutions in which a sample spreads its weight indiscriminately across many anchors. Together, these four components—anchor-based reconstruction, locality-aware initialization, distance-weighted regularization, and coefficient regularization—form a single coherent objective that learns a structured, nonnegative representation without ever forming the full sample-to-sample affinity matrix.

The mathematical machinery behind the approach draws on a rich lineage of research. Low-rank representation, introduced by Liu and colleagues, and sparse subspace clustering, developed by Elhamifar and Vidal, established the foundational idea that imposing structure on self-representation coefficients reveals subspace membership. Nonnegative matrix factorization, famously connected by Lee and Seung to the way humans learn parts of objects, supplies the theoretical justification for the nonnegative coefficients at the heart of LSSC. More recent work on anchor graphs and landmark-based spectral clustering demonstrated that working with a reduced set of representative points can preserve much of the quality of full spectral methods at a fraction of the cost. LSSC synthesizes these threads, but its distinctive contribution is the integration of structure-preserving regularization into the anchor-based pipeline, addressing the structural fidelity problem that earlier fast methods left unresolved.

To test the framework, the team ran extensive experiments on nine benchmark datasets, comparing LSSC against a battery of established clustering algorithms using four standard evaluation metrics: clustering accuracy (ACC), normalized mutual information (NMI), Purity, and the adjusted rand index (ARI). The results were striking. Among the compared methods that completed all nine datasets, LSSC achieved the best overall average rank across all four metrics, indicating that its advantage is not confined to one type of data or one evaluation criterion but reflects a consistently strong performance profile. On the large-scale datasets, where many competitors either slowed to a crawl or failed outright, LSSC maintained competitive clustering quality while demonstrating the favorable scalability that the anchor-based design was built to deliver.

The authors are careful to characterize the limits of their method as well as its strengths, a candor that lends the study particular scientific value. Their experiments reveal that LSSC shows sensitivity to highly imbalanced data distributions, a scenario in which some clusters contain vastly more samples than others. This vulnerability is a known hazard of anchor-based approaches, since anchors tend to be drawn disproportionately from dense regions of the data, leaving sparse clusters underrepresented in the reconstruction basis. By documenting this weakness explicitly, the researchers provide a clear empirical map of where the framework should be deployed and where practitioners should exercise caution, an honest assessment that is all too rare in a field often driven by headline benchmark numbers.

The implications reach well beyond the machine learning community. Subspace clustering underpins applications from image segmentation and hyperspectral image analysis to single-cell RNA sequencing, where biologists attempt to sort thousands of individual cells into functional types based on high-dimensional gene expression. A method that scales gracefully while preserving structural fidelity could make such analyses feasible on datasets that are currently out of reach, and the same logic applies to speech processing, anomaly detection, and any domain where high-dimensional observations hide low-dimensional group structure. The work, supported by research grants from Zhejiang Province and Hangzhou Dianzi University, arrives at a moment when the volume of unlabeled data is exploding faster than the tools to organize it. By showing that structure and scale need not be enemies, LSSC offers a template for the next generation of clustering algorithms: representations that are compact by design, yet rich enough to remember the geometry that made the clusters real in the first place.

Subject of Research: Scalable anchor-based subspace clustering for large high-dimensional datasets

Article Title: Large-scale structured subspace clustering

Article References: Chen, M., Yin, X., Huang, Q., Ding, J., Zhang, G., & Miao, X. (2026). Large-scale structured subspace clustering. Data Mining and Knowledge Discovery, 40(6), Article 99. https://doi.org/10.1007/s10618-026-01268-x

Image Credits: AI Generated

DOI: 10.1007/s10618-026-01268-x

Keywords: subspace clustering, anchor-based modeling, representation learning, affinity graph, manifold regularization, nonnegative matrix factorization, data mining, unsupervised learning, scalability, high-dimensional data, clustering algorithms, machine learning

The core insight behind EPR-C3 is that the search for the best subset should not be blind. Traditional approaches to the problem fall into two broad camps, each with well-known weaknesses. Stepwise procedures, which add or remove predictors one at a time based on statistical criteria, are fast but notoriously unstable, often missing the globally best model and producing results that shift with small perturbations of the data. Penalized regression methods such as the lasso, ridge, and elastic net shrink coefficients to perform implicit selection, but they transform the problem itself, replacing the ordinary least squares formulation with a modified objective and yielding biased estimates that complicate interpretation. Genetic algorithms and other stochastic search techniques can explore large spaces, but their reliance on random initialization and random operators means that two runs of the same algorithm on the same data can produce different answers, undermining the reproducibility that scientific work demands.

EPR-C3 takes a different path. The method is a deterministic multi-start neighborhood-search heuristic, meaning that it systematically explores multiple regions of the subset space using structured moves rather than random jumps, and that repeated runs on identical data produce identical results. The name encodes its workflow: expansion, perturbation, and reduction form the exploratory core, while C3 refers to a three-part constraint refinement stage comprising correlation cleanup, replacement recovery, and variance inflation factor pruning. Each stage addresses a specific pathology of high-dimensional regression. Expansion grows a candidate subset by adding predictors that promise improvement. Perturbation shakes up the current solution to escape local optima, the traps where a greedy search would otherwise stall. Reduction trims away predictors that no longer earn their place.

The C3 refinement stage is where the method embeds statistical admissibility directly into the search. Correlation cleanup removes predictors that are so strongly correlated with others already in the model that they add redundancy rather than information. Replacement recovery attempts to swap problematic predictors for alternatives that preserve predictive power while restoring stability. Variance inflation factor pruning, the final safeguard, quantifies how much the variance of an estimated coefficient is inflated by multicollinearity and eliminates predictors whose presence destabilizes the model. By enforcing these constraints during the search rather than checking them afterward, EPR-C3 ensures that every candidate model it evaluates is statistically admissible, avoiding the embarrassing situation in which a search returns a model with excellent fit but coefficients so unstable that they are meaningless.

A crucial design decision sets EPR-C3 apart from penalized approaches: it preserves the native ordinary least squares formulation throughout. The final output is not a shrunken or regularized set of coefficients but an explicit regression equation of the familiar form, with interpretable coefficients estimated by classical least squares. This matters enormously for applied researchers in fields such as chemistry, medicine, and the social sciences, where the goal is often not just prediction but understanding, and where an explicit equation with transparent coefficients is the deliverable that matters. The author’s own background in chemical modeling, including prior work on predicting acid dissociation constants in nitrogen compounds, reflects this applied orientation toward interpretable equations.

The benchmark results are striking. In analyses comparing EPR-C3 against exhaustive all-subsets regression, the heuristic recovered a high proportion of the Top-K admissible subsets, the best-ranked models that exhaustive search identifies, while dramatically reducing computational cost in highly combinatorial problems. In other words, when the exhaustive answer is computable, EPR-C3 usually finds it; when exhaustive search is impossible, EPR-C3 still delivers high-quality admissible models in a fraction of the time. The study goes further and proposes an empirical utility threshold, a practical guideline that tells analysts when the heuristic search becomes preferable to exhaustive enumeration, giving practitioners a principled rule for choosing between the two strategies rather than guessing.

The comparative evaluation is unusually thorough for a methods paper. EPR-C3 was benchmarked against stepwise procedures, penalized-regression preselection, filter and wrapper screening approaches, genetic algorithms, and branch-and-bound subset selection, the latter being the classical exact method that prunes the search space using bounds on the objective. Across these comparisons, a consistent pattern emerged: embedding admissibility constraints within the search workflow improves the tradeoff between recovery of the best models and computational efficiency. Competing methods that search first and check statistical validity later, or that abandon the least squares framework entirely, tend to either waste effort on inadmissible candidates or return models that require post-hoc repair. EPR-C3’s constraint-aware architecture avoids both failure modes.

Reproducibility and robustness received dedicated attention. Bootstrap analyses, in which the method was run on resampled versions of the data, showed close agreement with exhaustive selection in best-subset recovery, out-of-bag performance, and predictor-inclusion concentration, the degree to which the same predictors are selected consistently across resamples. This concentration measure is particularly important because instability of variable selection is one of the most cited criticisms of heuristic and stepwise methods. A seed-package analysis further confirmed that the deterministic nature of the algorithm holds in practice, with repeated runs converging on the same solutions. An applied case study demonstrated that EPR-C3 could reproduce a previously published multiple linear regression equation, recovering the same model with reduced runtime, providing a concrete demonstration that the method delivers on real problems rather than only on synthetic benchmarks.

The significance of this work extends beyond the specific algorithm. Variable selection remains one of the most contested practices in applied statistics, with methodologists repeatedly warning that automated selection can produce misleading inference, inflated significance, and models that fail to replicate. By making the search deterministic, constraint-aware, and anchored in the classical least squares framework, EPR-C3 offers a middle path between the computational impossibility of exhaustive search and the statistical compromises of penalization and greedy stepwise selection. The method is implemented in the author-maintained MLR-X project and publicly available, lowering the barrier for adoption. As datasets in genomics, chemometrics, economics, and the behavioral sciences continue to swell in dimensionality, tools that combine computational tractability with statistical admissibility and exact reproducibility are likely to find eager audiences. The study, published as volume 22, article 311 of the journal and supported by ANID-FONDECYT grant funding, suggests that the future of high-dimensional model selection may lie not in abandoning classical regression but in searching its vast model space more intelligently.

Subject of Research: Deterministic constraint-aware subset selection in high-dimensional multiple linear regression

Article Title: Epr-c3: a deterministic constraint-aware heuristic for high-dimensional subset selection in multiple linear regression

Article References: Alcázar, J. J. (2026). Epr-c3: a deterministic constraint-aware heuristic for high-dimensional subset selection in multiple linear regression. International Journal of Data Science and Analytics, 22(1), Article 311. https://doi.org/10.1007/s41060-026-01298-0

Image Credits: AI Generated

DOI: 10.1007/s41060-026-01298-0

Keywords: subset selection, multiple linear regression, ordinary least squares, multicollinearity, variance inflation factor, heuristic optimization, best-subset regression, statistical computing, high-dimensional data, model selection, deterministic algorithms, bootstrap validation

The Internet of Things presents security analysts with an unusually hostile data environment. Unlike traditional enterprise networks, IoT ecosystems mix smart thermostats, cameras, medical sensors, and industrial controllers, each producing traffic with distinct statistical fingerprints. The resulting datasets are high-dimensional, often containing hundreds of numerical and categorical features, and severely imbalanced, with benign traffic dwarfing attack traffic by orders of magnitude. Attacks themselves range from botnet command-and-control traffic to application-layer exploits, so the signals that betray an intrusion vary enormously in character. Machine learning classifiers trained on such data struggle not because the information is absent but because it is buried under redundant, irrelevant, and noisy variables that dilute the learning signal.

Feature selection is the discipline of separating signal from this noise. Rather than transforming features into new abstract dimensions, as feature extraction methods such as autoencoders do, feature selection identifies and retains the original attributes that carry the most discriminative power, preserving interpretability. Filter methods, which rank features using statistical measures such as mutual information and correlation without consulting any classifier, are prized for their speed and scalability. Their weakness is that any single ranking criterion has blind spots; a measure that captures linear relationships may miss nonlinear ones, and vice versa. Ensemble feature selection addresses this by aggregating the verdicts of several ranking algorithms, in the same spirit that ensemble classifiers combine many weak learners into one strong predictor.

EF²RB, developed by Chandam Chinglensana Singh, Nazrul Hoque, and Khumukcham Robindro Singh, extends this ensemble philosophy with a concept the authors call Ranking with Boosting. The framework begins by partitioning the feature space into manageable segments, a decision that proved essential for scalability. Each partition is then evaluated by multiple filter ranking algorithms, including methods built on mutual information such as mRMR, JoMIC, and MIFS-ND, alongside correlation-based ranking. A consensus voting mechanism then reconciles the individual rankings to produce a common feature subset, and a correlation-pruning stage removes redundant variables that carry overlapping information. An iterative refinement loop, functioning as the ranker booster, progressively sharpens the subset across rounds, analogous to how boosting algorithms iteratively focus on the hardest examples.

The ablation experiments conducted by the team offer a revealing anatomy of the framework. Using the RT-IoT2022 dataset alongside three non-IoT benchmarks from the UCI Repository, namely Mice Protein Expression, Spambase, and Optical Recognition of Handwritten Digits, the researchers disabled each component in turn while holding the rest constant. The consensus mechanism emerged as the single most influential element: switching it off on RT-IoT2022 shrank the selected feature set from 21 features to just 5 and dragged accuracy down from roughly 98.94 percent to 94.77 percent. Correlation pruning, by contrast, mainly trims redundancy, expanding the retained features from 21 to 23 when disabled while leaving accuracy almost unchanged.

The most striking ablation result concerned feature-space partitioning. When the researchers allowed the ranking algorithms to process the entire feature space at once, runtimes exploded to extraordinary levels, with mRMR requiring more than 20,500 seconds, JoMIC nearly 50,000 seconds, and MIFS-ND more than 9,600 seconds on a single dataset. Worse, the consensus mechanism failed to produce any common feature subset at all, which prevented the classification stage from running. On the handwritten digits dataset, disabling partitioning likewise halted the pipeline entirely. These findings underline that for genuinely high-dimensional data, how you divide the problem can matter as much as the algorithms you apply to it. Iterative feature selection, interestingly, contributed only marginal gains, suggesting the framework remains robust even without that refinement layer.

Beyond the ablation study, the team embedded EF²RB in a full IoT intrusion detection system and tested it with baseline machine learning classifiers on high-dimensional IoT intrusion datasets that include Bot-IoT, Edge-IIoTset, NSL-KDD, UNSW-NB15, N-BaIoT, and Mu-IoT, alongside classical benchmarks. The resulting detector consistently achieved high performance, with accuracy between 95 and 100 percent on certain datasets, even though the classifier worked from a dramatically reduced feature subset. Because filter methods dominate the framework, the heavy computation happens once, offline, before deployment, leaving the live detector fast enough for resource-constrained IoT gateways that cannot afford heavyweight deep learning inference.

The comparison with existing approaches is instructive. Prior ensemble methods such as IDS-EFS and rank aggregation schemes in software defect prediction have shown that combining filters improves stability, while metaheuristic wrappers guided by multiple rankers have pushed accuracy on imbalanced data. EF²RB distinguishes itself by uniting partitioning, multi-ranker consensus, redundancy pruning, and iterative boosting in one pipeline, and by demonstrating each component’s contribution through systematic ablation rather than reporting only aggregate results. The authors have also released their implementation publicly on GitHub, which lowers the barrier for security teams and researchers to reproduce, audit, and adapt the framework for their own network environments.

The broader significance lies in what this means for defending the rapidly expanding IoT attack surface. Botnet campaigns such as Mirai demonstrated years ago that insecure connected devices can be weaponized at internet scale, and detection systems have struggled to keep pace with both the volume and heterogeneity of device traffic. A feature selection method that is dataset-agnostic, as the UCI benchmark results indicate, offers security engineers a reusable tool rather than a bespoke fix: the same pipeline that compresses IoT intrusion data can also condense spam indicators or biomedical measurements. As regulatory pressure and liability concerns push manufacturers toward hardened devices, methods like EF²RB supply the monitoring side of that equation, turning enormous, noisy traffic streams into compact feature sets that lightweight classifiers can act on in near real time.

Limitations remain, and the authors are candid about them. The evaluation relied on established benchmark datasets rather than newly generated traffic, so real-world deployment performance will depend on how faithfully those datasets mirror live networks, which constantly evolve as attackers adapt. The framework’s runtime, though far better with partitioning, still depends on computationally expensive mutual information estimates, which may need further optimization for streaming or federated deployments. Yet the core result stands: disciplined consensus among diverse ranking perspectives, guided by a boosting-inspired refinement, can extract a small, potent feature subset from the chaos of high-dimensional IoT data. In a field where every percentage point of detection accuracy translates into intercepted attacks, that is a consequential advance.

Subject of Research: Ensemble filter feature selection with ranker boosting for high-dimensional IoT intrusion detection

Article Title: EF(^{2})RB: Ensemble of filter feature selection methods with ranker booster for classification of high-dimensional IoT intrusion data

Article References: Chinglensana Singh, C., Hoque, N., & Robindro Singh, K. (2026). EF$$^{2}$$RB: Ensemble of filter feature selection methods with ranker booster for classification of high-dimensional IoT intrusion data. Knowledge and Information Systems, 68(1), Article 252. https://doi.org/10.1007/s10115-026-02871-6

Image Credits: AI Generated

DOI: 10.1007/s10115-026-02871-6

Keywords: feature selection, feature ranking, dimensionality reduction, ensemble learning, IoT security, intrusion detection, machine learning, mutual information, high-dimensional data, class imbalance, cybersecurity, ranking with boosting