Graph embedding, the task the model targets, sits at the heart of many modern data applications. Social networks, citation networks, knowledge graphs, protein interaction networks and recommendation systems all share a common mathematical structure: entities are nodes, relationships are edges, and nodes often carry rich attribute information such as text or numerical features. Graph embedding compresses this high-dimensional, sparse structure into compact low-dimensional vectors that downstream algorithms can consume. The quality of those vectors determines how well a system can predict missing links, classify nodes into categories, or cluster similar entities together. As graphs grow to millions of nodes, the efficiency of the embedding method becomes as important as its accuracy.

The dominant tools for this job are Graph Autoencoders, known as GAEs, which combine graph convolution operations with an encoder-decoder framework. Graph convolutions allow each node to aggregate information from its neighbors, weaving structural and attribute information into a single representation. But these models are trained with gradient descent, an iterative optimization procedure in which parameters are nudged repeatedly in the direction that reduces reconstruction error. The authors of the new study point out two persistent problems with this approach: convergence can be slow, particularly on large graphs, and the optimization landscape is riddled with local optima, meaning the model can settle into a mediocre solution that it cannot escape. Both issues translate into wasted computation and unpredictable quality.

RCGELM-AE takes a fundamentally different route by building on the Extreme Learning Machine Autoencoder, or ELM-AE, a family of models in which the input-to-hidden weights are randomly assigned and fixed, and the output weights are solved in closed form using linear algebra rather than iterative optimization. Because the hidden layer parameters never need to be tuned, training reduces to a single matrix computation, which is fast, stable and immune to the local optima that plague gradient descent. The generalized variant, GELM-AE, extends this idea with a regularization term that improves generalization. What plain GELM-AE lacks, however, is any notion of graph topology. It treats each node as an independent data point, blind to the edges that define the network. The new model closes that gap with two carefully engineered components.

The first component embeds graph convolution operations directly between the input and hidden layers of the autoencoder. In practical terms, before the random projection takes place, each node’s features are smoothed over its local neighborhood, so the representation that enters the hidden layer already carries information about who a node is connected to. This single, strategically placed convolution compensates for the inherent inability of GELM-AE to capture topological information, allowing the model to integrate structure and attributes in one pass without the deep stacks of propagation layers used in conventional graph neural networks.

The second component is a residual compensation mechanism, an idea inspired by the residual learning revolution in computer vision. Instead of stacking many graph propagation layers to extract deeper features, which is known to cause over-smoothing, a pathology in which node representations become indistinguishable as information is averaged over and over, RCGELM-AE feeds the reconstruction errors of the encoder back into the pipeline. These errors, the parts of the input the first encoding pass failed to capture, are treated as a new signal and encoded again, extracting deeper-level features layer by layer. Each compensation stage therefore adds representational depth without adding repeated graph smoothing, sidestepping the over-smoothing risk that limits the depth of conventional deep graph models while boosting the model’s expressive power.

The empirical results reported in the study are substantial. On link prediction tasks, where the goal is to infer missing or future connections between nodes, the model improved the area under the receiver operating characteristic curve, AUC, by 4.11 to 8.93 percent and average precision, AP, by 4.01 to 10.01 percent relative to the baselines. On node classification tasks, where each node must be assigned to a category, the F1-micro and F1-macro scores rose by 0.42 to 4.97 percent and 0.48 to 5.27 percent respectively. Crucially, these gains came with a notable reduction in running time, because the closed-form learning scheme replaces hundreds or thousands of gradient updates with a single solve.

The scalability evidence is perhaps the most compelling part of the study. The authors evaluated RCGELM-AE on two large benchmark graphs from the Open Graph Benchmark, ogbn-arxiv and ogbl-ppa, which contain substantially more nodes and edges than the datasets used in the main experiments. Running on a 24 GB NVIDIA GeForce RTX 3090 GPU, the model achieved the best AUC and AP scores on both datasets against representative gradient-trained baselines including GAE, VGAE, LGAE and DGNN. The total training times were remarkable: just 2.12 seconds on ogbn-arxiv and 9.53 seconds on ogbl-ppa. The baselines, by contrast, require repeated epoch-wise optimization, with each epoch costing a full pass of forward and backward computation. For iterative models the reported figure is only the average time per epoch, meaning their total cost is orders of magnitude higher.

The implications extend beyond raw speed. Gradient-free training eliminates a whole class of engineering headaches: learning-rate schedules, initialization sensitivity, early-stopping heuristics and the variance introduced by stochastic mini-batching all become irrelevant when the solution is computed analytically. That stability is attractive for applications where reliability matters, such as fraud detection on financial networks, drug discovery on molecular graphs, or knowledge graph completion in large-scale information systems. It also lowers the barrier for researchers and organizations without access to extensive GPU clusters, since a model that trains in seconds on a single consumer graphics card democratizes access to state-of-the-art graph learning.

The work, led by Xinyi Lin, Xiaoyun Chen, Shulan Zheng and Wenjian Chen of the College of Mathematics and Statistics at Fuzhou University, and supported by the Natural Science Foundation of Fujian Province, does not claim that gradient descent is obsolete. Random projection methods have their own trade-offs, and the fixed random layer must be wide enough to capture the relevant feature space. But the study makes a persuasive case that the deep learning community’s default assumption, that competitive graph representations require iterative optimization, deserves scrutiny. As graphs in science, commerce and social media continue to swell toward planetary scale, a framework that delivers better accuracy in a fraction of the time may prove less a curiosity and more a glimpse of where efficient machine learning is headed.

Subject of Research: A gradient-descent-free graph embedding autoencoder combining graph convolution and residual compensation for efficient link prediction and node classification

Article Title: RCGELM-AE: an efficient graph embedding deep model without gradient descent

Article References: Lin, X., Chen, X., Zheng, S., & Chen, W. (2026). RCGELM-AE: an efficient graph embedding deep model without gradient descent. Applied Intelligence, 56(14), Article 404. https://doi.org/10.1007/s10489-026-07437-1

Image Credits: AI Generated

DOI: 10.1007/s10489-026-07437-1

Keywords: graph embedding, extreme learning machine, autoencoder, gradient descent, graph convolutional networks, link prediction, node classification, residual compensation, over-smoothing, Open Graph Benchmark, deep learning, Applied Intelligence

The research, published in the journal Plant Methods, tackles a deceptively simple question: how do you make irregularly shaped objects like maize ears line up and flow smoothly along a vibrating conveyor? Unlike spheres or cubes, maize ears tumble, rotate, and jam unpredictably. Traditional response surface methodology, the statistical workhorse engineers typically use to tune machinery parameters, simply cannot capture the strongly nonlinear behavior of these tumbling biological objects. The team, led by co-first authors Yingjie Li and Yalei Xu, recognized that a fundamentally new approach was needed.

Their solution was a three-part framework they describe as mechanism simulation, intelligent modeling, and global optimization. The first step involved building a discrete element model, a computational technique that treats each maize ear as an individual particle with realistic physical properties, allowing researchers to simulate exactly how thousands of ears would behave under different vibration conditions. This digital twin of the conveying process gave them a window into physics that would be nearly impossible to observe directly at speed.

To train their models, the researchers needed high-quality data, and lots of it. They employed a hybrid sampling strategy that combined Box-Behnken designs with Latin hypercube sampling, generating 337 sets of high-fidelity simulation data. This clever pairing ensured that the data covered both the systematic experimental design points and the broad parameter space needed for machine learning. When they validated their simulations against physical bench experiments across multiple operating conditions, the average relative error came in at less than five percent, a remarkably tight agreement for a system involving chaotic biological materials.

With the dataset in hand, the team built three competing predictive models: the traditional response surface methodology, a neural network architecture called an extreme learning machine, and an enhanced version of that network tuned by a grey wolf optimizer. The grey wolf optimizer is a nature-inspired algorithm that mimics the leadership hierarchy and hunting strategy of wolf packs to search parameter space efficiently. Five-fold cross-validation revealed a decisive winner: the grey wolf-optimized extreme learning machine achieved a coefficient of determination exceeding 0.98, dramatically outperforming the classical statistical approach on this stubbornly nonlinear problem.

The optimized model then pointed the way to an ideal operating regime: a vibration frequency of 16 hertz, an amplitude of 6 millimeters, and a vibration direction angle of 30 degrees. Under these conditions, the model predicted a feeding rate of 0.9981 ears per second, meaning the conveyor would deliver essentially one ear every second without interruption. The coefficient of variation, a measure of flow uniformity, was predicted at 43.59 percent. When the team put these parameters to the test on a physical three-level bench system, reality matched prediction almost exactly, with a measured feeding rate of 1.0056 ears per second and a coefficient of variation of 43.92 percent.

Perhaps most striking was the comparison with the conventional approach. Compared with the optimization results from response surface methodology alone, the new method increased the feeding rate by 41.6 percent while cutting the coefficient of variation by 29.6 percent. In practical terms, that means more ears flowing through the system per second, arriving in a steadier, more predictable stream, and with far less of the jostling and impact that causes mechanical damage to seed samples.

Beyond the headline numbers, the study delivered something arguably more valuable: a mechanistic understanding of why certain vibration settings work better than others. The researchers found that vibration frequency is the dominant factor controlling how quickly maize ears converge to an ordered posture, and therefore how fast the system can feed. Amplitude and vibration direction angle, by contrast, act as a coupled pair that governs the smoothness of posture adjustment, and thus the stability of the conveying process. This division of labor gives engineers a clear design principle: tune frequency for speed, then balance amplitude and angle for stability.

The implications extend well beyond maize. The framework of mechanism simulation, intelligent modeling, and global optimization is explicitly designed to be transferable to other irregularly shaped seeds and biological materials that resist conventional handling. In plant breeding programs, where seed samples are irreplaceable and mechanical damage can destroy years of genetic work, the ability to convey materials gently and in an orderly fashion directly protects research value. The method could also inform the design of seed counters, singulators, and precision planters that depend on reliable single-file feeding.

The work, supported by the Xinjiang Uygur Autonomous Region High-end Intelligent Agricultural Machinery Industry Innovation Research Institute, represents a growing trend in agricultural engineering: replacing trial-and-error tuning with simulation-driven, machine-learning-accelerated design. As farms and breeding facilities demand ever-faster, gentler handling of biological materials, approaches like this one, which marry the physics of granular flow with the pattern-finding power of modern algorithms, are likely to become the standard rather than the exception. For the humble maize ear, it means a smoother ride from field to laboratory, and for the scientists who depend on those ears, it means better data with less loss.

Subject of Research: Ordered vibration conveying mechanism and parameter optimization for maize ears in plant breeding

Article Title: Mechanism and parameter optimization method for ordered vibration conveying of maize ears

Article References: Li, Y., Xu, Y., Li, J., Li, Y., & Wang, X. (2026). Mechanism and parameter optimization method for ordered vibration conveying of maize ears. Plant Methods. https://doi.org/10.1186/s13007-026-01580-z

Image Credits: AI Generated

DOI: 10.1186/s13007-026-01580-z

Keywords: maize ears, vibration conveying, discrete element method, extreme learning machine, grey wolf optimizer, parameter optimization, plant breeding, seed handling, agricultural machinery, surrogate modeling, response surface methodology, ordered conveying

The work, carried out by Ghabriel A. Gomes de Sá, Cristiano Hora Fontes and Marcelo Embiruçu at the Federal University of Bahia, addresses a blind spot in the rapidly growing field of physics-informed machine learning. The dominant framework, Physics-Informed Neural Networks, or PINNs, embeds physical knowledge by adding the deviation from a known phenomenological model, typically differential equations derived from conservation principles, as a penalty term in the loss function. This strategy has powered applications as diverse as heat conduction in porous media, hydraulic fracturing assessment, structural stability prediction and cement hydration modeling. But all of these applications share a crucial presupposition: somebody has already derived the governing equations. In most real-world data-driven modeling problems, no such model exists, either because the physics is incompletely understood or because unquantifiable disturbances dominate the process.

The researchers argue that even in this model-free regime, there is almost always some minimal physical knowledge worth enforcing: the sign of the static gain between each input and the output. The static gain is the ratio of the change in output to the change in a given input at a reference steady state, and its sign encodes the direction of the causal effect. Does raising cycle time lower processor performance? Does increasing lipophilicity raise aquatic toxicity? These are qualitative facts that domain experts, established theory or quantitative structure-activity relationships can supply even when full equations cannot. The team’s central claim is that this elementary piece of knowledge should be treated not as a soft preference but as a hard constraint, guaranteed to hold in the final model rather than merely encouraged.

Formally, the approach, which the authors call Gain-Constrained Training, sets up a constrained optimization problem. The loss function remains the familiar mean squared error between predicted and measured outputs, but the training is now subject to nonlinear inequality constraints requiring each predicted static gain to carry its expected sign. The gains are evaluated numerically by finite differences at a reference point, the mean of the test inputs, with a small margin ensuring strict inequality. Because the case studies involve steady-state problems where the effect of each input is monotonic across the operating region, enforcing the sign at a single representative point suffices. The formulation also generalizes naturally: if more information is available, the constraints can be tightened into ranges on the gains, or even into relationships linking gains to other variables, moving the method closer to a fully structured model of the phenomenon.

Enforcing such constraints raises an obvious practical question: how do you actually train a network this way? The researchers tested three distinct learning regimes to make the answer as general as possible. Two were gradient-based methods with different philosophies of weight initialization: WILCAR, a constructive algorithm that linearizes a single-neuron network to obtain analytical initial weights and then grows the hidden layer incrementally, reusing previously learned weights; and RIXM, a conventional random initialization following the Xavier scheme, in which every training run starts from scratch. The third was the Extreme Learning Machine, a gradient-free method in which hidden-layer weights are fixed at random values and the output weights are computed analytically via the Moore-Penrose pseudo-inverse, making it dramatically faster than iterative training. In every constrained variant, the Sequential Least Squares Programming algorithm minimizes the loss subject to the gain-sign constraints, and a restart strategy reinitializes the weights, up to one hundred attempts, whenever the resulting model fails to satisfy every prescribed sign.

A key insight of the paper concerns where the optimization begins. Weight initialization is conditioned on the predefined gain signs so that, whenever possible, training starts inside the feasible region, the subset of parameter space in which all constraints are already satisfied. In the constructive WILCAR method, this conditioning is strengthened by applying the simplex algorithm to initialize the weights of newly added hidden units. Starting within the feasible region improves convergence and reduces the number of restarts needed to achieve full sign conformity, turning what could be an expensive combinatorial burden into a manageable computational overhead.

The case studies drew on three benchmark datasets from the UCI Machine Learning Repository, chosen because literature and expert knowledge support clear expectations about causal directions. The Computer Hardware dataset, with 209 machine configurations from the early 1980s, asks networks to predict published relative performance from cycle time, memory limits, cache size and I/O channel counts; architecture dictates that cycle time should carry a negative gain while memory, cache and channel variables carry positive gains. The two toxicology benchmarks are harder. The QSAR Fish Toxicity dataset contains 908 compounds with six molecular descriptors predicting the median lethal concentration in fathead minnows, while the QSAR Aquatic Toxicity dataset comprises 546 molecules with eight descriptors predicting 48-hour toxicity in the crustacean Daphnia magna. Here the expected signs derive from structure-activity relationships: lipophilicity increases baseline narcotic toxicity, electrophilic descriptors raise reactivity with biological nucleophiles, and polar surface descriptors behave in more mechanism-dependent ways.

The results deliver a striking verdict on unconstrained training. Across all three datasets and all three learning methods, models that were properly trained, validated and selected by five-fold cross-validation nevertheless failed to reproduce the expected gain signs. On the fish toxicity problem, for example, the unconstrained Extreme Learning Machine achieved full sign conformity on only 73.3 percent of inputs, and on the more complex aquatic toxicity dataset none of the unconstrained methods reached full conformity at all, with conformity apparently degrading as problem complexity increased. In other words, good regression metrics such as low root-mean-squared error and high coefficient of determination provide no assurance that a black-box model is even qualitatively consistent with the physics of the situation.

Against this baseline, the constrained methods performed flawlessly on the criterion that matters most: every constrained variant achieved a 100 percent Signal Conformity Rate across all three datasets, confirming the effectiveness of the constrained optimization and restart strategy. Just as importantly, this qualitative guarantee came essentially for free. On the Computer Hardware problem, average test R-squared values ranged from 0.856 to 0.902 across all six approaches, with constrained and unconstrained distributions nearly indistinguishable, and ELM-based methods performing best. On the fish toxicity data, constrained models showed slightly lower average R-squared values, a modest price for guaranteed physical consistency. On the aquatic toxicity dataset, the constrained RIXM actually outperformed its unconstrained counterpart while using dramatically fewer neurons.

Perhaps the most intriguing finding is that constraints made the networks smaller. Learning curves showed that unconstrained methods often needed hundreds of hidden units to reach their best performance, whereas constrained methods achieved near-optimal accuracy with networks of fewer than fifty units, and in one case as few as fifteen. The authors interpret this through the lens of effective capacity: by restricting the parameter search to solutions consistent with the prescribed qualitative behavior, the constraints act as an inductive bias that reduces the complexity of the model class, curbing the familiar tendency of unconstrained networks to grow in order to fit noise. This echoes the principle of structural risk minimization and suggests the constraints function as a form of regularization far more meaningful than generic weight decay. Computational costs told a similarly favorable story: constrained gradient-based training generally required less effort than its unconstrained counterpart, and although constraining ELM added time, it remained faster than any gradient-based method.

The broader implications reach well beyond these three benchmarks. The authors emphasize that the framework is generic, applicable to shallow and deep architectures, static and recurrent networks alike, and extensible to range constraints and richer relationships among gains whenever more physical knowledge is available. For practitioners in engineering, cheminformatics, ecotoxicology and process modeling, the message is direct: a validated model with excellent metrics can still be qualitatively wrong, and the fix requires neither governing equations nor penalty-weight tuning, only the often readily available knowledge of which way each input pushes the output. In an era when machine learning models increasingly inform decisions about safety, health and design, ensuring that a network at least gets the sign of causality right may be the minimum standard of physical honesty that any deployed model should meet.

Subject of Research: Physics-informed training of feedforward neural networks with hard static gain-sign constraints for regression problems without phenomenological models

Article Title: Physics-informed neural networks without a phenomenological model available – applications in regression problems

Article References: de Sá, G. A. G., Fontes, C. H., & Embiruçu, M. (2026). Physics-informed neural networks without a phenomenological model available – applications in regression problems. Neural Computing and Applications, 38(17), Article 726. https://doi.org/10.1007/s00521-026-12462-9

Image Credits: AI Generated

DOI: 10.1007/s00521-026-12462-9

Keywords: physics-informed neural networks, machine learning, regression, static gain constraints, feedforward neural networks, Extreme Learning Machine, weight initialization, QSAR toxicity, model regularization, inductive bias, UCI benchmarks, qualitative consistency

The research, conducted by Furquan Ahmad of the National Institute of Technology Patna in India and Warit Wipulanusat of Thammasat University in Thailand, addresses a long-standing challenge in geotechnical engineering. Predicting the factor of safety against liquefaction requires untangling complex nonlinear interactions among soil properties, seismic loading conditions, and in-situ measurements. Traditional simplified procedures, developed over decades from case histories of past earthquakes, rely on empirical charts and hand calculations that can miss subtle parameter interactions. Machine learning offers a way forward, but conventional neural networks are often criticized as opaque black boxes, and their training can be slow and unstable. The new framework tackles both problems at once, pairing a fast-learning network architecture with a transparent layer of explainability analysis.

At the heart of the framework is the Extreme Learning Machine, or ELM, a type of single-hidden-layer feedforward neural network introduced in the mid-2000s. Unlike conventional neural networks that iteratively adjust all their internal weights through backpropagation, an ELM assigns its hidden-layer weights randomly and then solves for the output weights analytically in a single step. This makes training dramatically faster while often achieving excellent generalization. However, the random initialization of hidden-layer parameters can leave performance on the table. The researchers solved this by coupling the ELM with four metaheuristic optimization algorithms, each inspired by different natural processes: the Adaptive Opposition Slime Mould Algorithm, Harris Hawks Optimization, Particle Swarm Optimization, and the Reptile Search Algorithm. These algorithms search the space of possible network parameters to find configurations that minimize prediction error, effectively tuning the network far beyond what random initialization alone could achieve.

Each optimizer brings a distinct search philosophy. Particle Swarm Optimization, one of the oldest and most widely used metaheuristics, mimics the social behavior of bird flocks, with candidate solutions sharing information about promising regions of the search space. Harris Hawks Optimization models the cooperative hunting strategies of hawks, dynamically switching between exploration and exploitation phases to chase down optimal solutions. The Reptile Search Algorithm draws on the hunting and social behaviors of crocodiles, while the Adaptive Opposition Slime Mould Algorithm simulates the oscillating foraging patterns of slime moulds, enhanced with an opposition-based learning mechanism that evaluates candidate solutions and their mirrored counterparts to accelerate convergence. By benchmarking all four against the same ELM backbone and the same dataset, the study provides a fair head-to-head comparison of how each optimizer shapes predictive performance.

Before any model training began, the researchers applied a disciplined feature screening process grounded in information theory. Using a database of 309 case histories based on cone penetration test, or CPT, measurements, they employed Mutual Information Analysis and Shannon Entropy Analysis to quantify how much information each candidate input variable carries about the liquefaction outcome. Mutual information measures the reduction in uncertainty about one variable given knowledge of another, while Shannon entropy, a concept borrowed from communication theory, quantifies the inherent uncertainty or information content of a variable. This pre-processing step ensures that the models focus on genuinely informative features rather than redundant or noisy ones, a critical consideration when working with real-world geotechnical data that often contains correlated measurements.

The performance results are striking. The best model, ELM-AOSMA, achieved coefficients of determination of 0.9912 on the training dataset and 0.9900 on the testing dataset, with root mean square errors of just 0.0146 and 0.0160 respectively. In practical terms, an R-squared value approaching 0.99 means the model explains nearly all of the variance in the liquefaction factor of safety across the case database, leaving only a tiny fraction unexplained. The models were validated through 5-fold cross-validation, a robust technique in which the data is partitioned into five subsets and the model is repeatedly trained and tested on different combinations to guard against overfitting. Beyond the headline metrics, the evaluation battery included the Nash-Sutcliffe coefficient, weighted mean absolute percentage error, and the A20-index, which measures the proportion of predictions falling within twenty percent of observed values.

The study also went beyond standard statistical scores with a suite of diagnostic analyses. Regression Error Characteristic curves were used to visualize the trade-off between prediction tolerance and the fraction of samples accurately predicted, while rank analysis and curve fitting analysis compared the relative standing of the four hybrid models. Convergence analysis tracked how quickly each optimizer approached its best solution, and computational efficiency was measured directly. ELM-AOSMA converged within approximately 400 iterations while requiring the lowest computational time of the four models, at 220.68 seconds. For engineers who may need to run liquefaction assessments across many sites or within probabilistic reliability frameworks that demand thousands of repeated evaluations, this combination of accuracy and speed is a meaningful advantage over slower, more computationally demanding alternatives.

Perhaps the most consequential contribution is the framework’s commitment to explainability. Using SHapley Additive exPlanations, a technique from cooperative game theory that allocates credit for each prediction to individual input features, and Partial Dependence with Individual Conditional Expectation analysis, which traces how predictions change as each variable varies, the researchers identified the dominant drivers of liquefaction behavior. Four parameters emerged as the most influential: the effective vertical stress, denoted sigma prime v; the normalized and cleaned cone penetration tip resistance, qc1Ncs; the total vertical stress, sigma v; and the magnitude scaling factor, MSF, which adjusts cyclic stress ratios for earthquakes of different magnitudes. This hierarchy aligns with established geotechnical understanding, since liquefaction resistance fundamentally depends on how tightly soil particles are confined and how densely they are packed, but the explainable AI analysis quantifies these relationships in a way empirical charts cannot.

The implications for practice extend to performance-based geotechnical design and reliability-oriented assessment. Modern seismic design codes increasingly demand not just a binary safe-or-unsafe verdict but a quantified factor of safety that can feed into probabilistic risk calculations. A model that is both highly accurate and computationally lightweight can serve as a surrogate within Monte Carlo simulations or first-order reliability methods, where the simplified empirical procedures would be evaluated tens of thousands of times. The interpretability layer also matters for professional acceptance: engineers are understandably reluctant to trust black-box predictions for safety-critical infrastructure, and SHAP-based attribution gives them a window into the reasoning behind each assessment, enabling sanity checks against engineering judgment.

The work arrives amid a broader wave of machine learning adoption in geotechnical engineering, with researchers applying hybrid models to slope stability, sheet pile walls, embankment performance in seismic zones, and liquefaction potential mapping in regions from the Indo-Gangetic plain to Kolkata. What distinguishes this study is its end-to-end rigor: information-theoretic feature selection, systematic optimizer benchmarking, cross-validated performance evaluation, and explainable AI diagnostics all integrated within a single framework. As earthquake-prone cities continue to grow and climate-driven changes alter groundwater conditions that govern soil saturation, tools that can rapidly and transparently assess liquefaction risk will become increasingly valuable. The ELM-AOSMA framework demonstrates that the next generation of geotechnical prediction tools need not force a choice between accuracy, speed, and interpretability; with careful design, they can deliver all three at once.

Subject of Research: Hybrid machine learning prediction of seismic soil liquefaction using cone penetration test data

Article Title: An information-theoretic and explainable hybrid extreme learning machine framework for seismic soil liquefaction assessment

Article References: Ahmad, F., & Wipulanusat, W. (2026). An information-theoretic and explainable hybrid extreme learning machine framework for seismic soil liquefaction assessment. Earth Science Informatics, 19(11), Article 195. https://doi.org/10.1007/s12145-026-02248-x

Image Credits: AI Generated

DOI: 10.1007/s12145-026-02248-x

Keywords: soil liquefaction, earthquake engineering, extreme learning machine, metaheuristic optimization, SHAP explainability, cone penetration test, mutual information, Shannon entropy, geotechnical engineering, machine learning, factor of safety, seismic hazard

The research, published in the International Journal of Machine Learning and Cybernetics by Shamama Anwar and Mayank Kumar of the Birla Institute of Technology, Mesra, addresses a problem that has quietly constrained drone-based traffic analytics for years. Conventional convolutional neural networks can classify vehicles with high accuracy, but training them requires iterative optimization of millions of weights through many passes over the data. That process is computationally intensive, energy hungry, and poorly matched to the embedded processors aboard small UAVs. The authors’ answer is to stop training the feature extractor altogether and let randomness do part of the work.

The core of the proposed framework is a Randomized Convolutional Neural Network, or RCNN. Instead of learning filters through backpropagation, the RCNN uses predefined random filters to transform raw images into rich feature representations. This idea draws on a broader body of research showing that randomness in neural networks, when properly harnessed, can serve as a powerful form of regularization and dramatically reduce training cost. Because the filters are fixed rather than learned, the expensive gradient-descent stage of feature learning disappears. What remains is a classification problem on top of those features, and this is where the architecture splits into two variants: one feeding the randomized features into an Extreme Learning Machine, the other into a Dense Neural Network.

The Extreme Learning Machine component is the more radical of the two. Introduced formally by Huang and colleagues in 2006, the ELM assigns random weights to its hidden layer and then solves for the output weights in a single linear step, reducing what is normally an iterative training process to an analytical calculation. The result is a classifier that trains orders of magnitude faster than a conventional neural network. Combined with the randomized convolutional front end, the full RCNN-ELM pipeline eliminates nearly all of the slow, iterative training that makes standard deep learning pipelines unsuitable for resource-limited platforms, while still producing discriminative features that separate cars, buses, trucks, and other vehicle categories.

Testing was deliberately harsh. The researchers evaluated the framework on two datasets designed to stress real-world deployment. The first, DAWN, is a benchmark specifically curated for vehicle detection in adverse weather, containing images shot through fog, rain, snow, and other visibility-degrading conditions that routinely defeat vision systems trained on sunny-day data. The second consists of UAV-based traffic images, capturing vehicles from the steep, oblique angles and variable altitudes that drones experience in flight. Both settings are representative of the conditions under which intelligent transportation systems must actually operate, and both are notoriously unforgiving to models tuned on idealized imagery.

The headline results favor the ELM variant. On the adverse-weather DAWN dataset, RCNN-ELM achieved a validation accuracy of 85.42 percent, outperforming the RCNN-DNN configuration, which reached 81.16 percent, while requiring substantially less training time. That combination of higher accuracy and lower cost is unusual; typically researchers trade one for the other. The authors backed the findings with statistical analysis across multiple runs, confirming that the performance differences were reliable rather than artifacts of a lucky random seed. Cross-dataset evaluation, in which a model trained on one data source was tested on the other, showed that the framework generalizes well across very different imaging conditions, a property critical for systems that cannot be retrained every time the weather or camera geometry changes.

Beyond the headline numbers, the study includes ablation studies that peel apart the contribution of each component, along with interpretability analyses that probe what the randomized features are actually detecting. These experiments serve a practical purpose. In safety-relevant applications such as traffic monitoring, understanding why a model classifies a blurred, rain-streaked blob as a truck rather than a car matters as much as the raw accuracy figure. The interpretability results, together with the ablations, support the central claim that randomized deep features capture genuinely useful structure in vehicle imagery even without task-specific training of the convolutional stage.

The work builds on a long line of vehicle classification research, from early SIFT- and HOG-based detectors paired with support vector machines, through dynamic Bayesian networks for aerial surveillance, to the modern YOLO family of real-time object detectors that now spans versions from YOLOv5 through YOLOv8 and beyond. Each generation has improved detection performance, but the computational appetite has generally grown in step. Recent surveys of drone-based road traffic monitoring highlight exactly this bottleneck: the platforms that would benefit most from autonomous vision are the ones least able to run state-of-the-art models at full strength. The randomized-hybrid approach offers an escape route that does not demand waiting for more powerful onboard hardware.

There is also a data-quality story embedded in the study. The framework’s robustness on DAWN speaks to a broader challenge in the field, where models trained on curated benchmarks frequently collapse when confronted with weather-induced image degradation. Researchers have attacked this problem with data augmentation, synthetic data generation, and weather-specific architectures, yet adverse conditions remain a leading cause of failure in deployed vision systems. Demonstrating strong accuracy directly on an adverse-weather benchmark, rather than on clean data with hoped-for transfer, gives the RCNN-ELM approach a credibility that many efficiency-focused papers lack.

The implications extend past traffic cameras. Randomized neural networks and extreme learning machines have found applications from large-scale scene recognition to lane detection, and the pattern they share, replacing slow iterative optimization with fast randomized feature generation plus analytical or lightweight learning, is attractive anywhere compute is scarce: satellites, medical devices at the edge, industrial sensors, and battery-powered robots. For urban planners and transportation agencies, the near-term significance is more concrete. The study suggests that drones patrolling highways in fog and rain could classify vehicles in real time on modest hardware, feeding accurate traffic data into intelligent transportation systems without the latency of offloading video to distant servers. The research was funded by the Department of Science and Technology of India under its Promotion of University Research and Scientific Excellence scheme, and the authors note that the RCNN-ELM framework offers an efficient and robust solution for real-time UAV-based traffic monitoring. If the efficiency gains hold up in field deployments, the humble combination of random filters and a fast linear classifier could become a quiet workhorse of the smart-city sensor layer.

Subject of Research: Lightweight hybrid deep learning for real-time vehicle classification in adverse weather and UAV-based traffic monitoring

Article Title: Hybrid randomized deep feature learning and extreme learning machines for vehicle classification under challenging visual conditions

Article References: Anwar, S., & Kumar, M. (2026). Hybrid randomized deep feature learning and extreme learning machines for vehicle classification under challenging visual conditions. International Journal of Machine Learning and Cybernetics, 17(9), Article 457. https://doi.org/10.1007/s13042-026-03296-x

Image Credits: AI Generated

DOI: 10.1007/s13042-026-03296-x

Keywords: vehicle classification, randomized neural networks, extreme learning machine, RCNN-ELM, adverse weather, UAV traffic monitoring, intelligent transportation systems, computer vision, deep learning, DAWN dataset, lightweight architecture, real-time detection