Tag: time series

Researchers at the University of KwaZulu-Natal have developed an algorithm that detects and repairs timestamp-displaced sensor data by exploiting spatial ...

A leakage-aware artificial intelligence framework trained on six seasons of commercial farm data forecasts daily strawberry yields with a pretrained ...

A new spiral-based transformation converts time series into images that let pretrained vision models outperform specialized anomaly detectors across 23 ...

Researchers at Anhui Agricultural University have developed DE-TabNet, a semi-supervised AI model that detects sensor faults in smart drip irrigation ...

Researchers in China have developed OzoneKBNet, a retrieval-augmented AI framework that mines historical pollution episodes to deliver more accurate 48-hour ...

A 13-year Baltic Sea time series shows that bacterial communities at identical temperatures differ radically between spring and autumn, revealing ...

A hybrid CNN–LSTM deep learning model tuned by Bayesian optimization outperformed classical and standalone neural benchmarks in quarterly GDP forecasts ...

Researchers in Beijing have unveiled a wavelet-enhanced model combining Mamba sequence modeling and graph neural networks that outperforms prevailing methods ...

A five-year study of nine commodities in Maharashtra found that ensemble machine learning models beat seasonal benchmarks but could not ...

Spanish researchers have developed a correlation-based k-medoids clustering method that classifies citrus price dynamics into three stable, economically meaningful patterns ...