Tag
The paper introduces HARN, a hierarchical associative resonance network for event-driven multi-timeframe forecasting in financial time series, showing competitive results against baselines through evaluations on multiple assets.
TimeEvo is a failure-driven self-evolution method for time series agents that diagnoses capability gaps, synthesizes tools, and improves accuracy across tasks and backbones.
This paper reviews the adaptation of stochastic differential equations with neural networks for astrophysical time series analysis, introduces Continuous-Delayed-Memory Stochastic Gradient Descent for improved optimization, and proposes a continuous-time reinforcement learning framework.
This paper proposes Fractional-Order Ordinary-Least-Squares Grid-Search (FO-GS), a method for identifying fractional-order linear time-invariant systems from a single trajectory, with theoretical error bounds and experimental validation.
This paper introduces Calendar-SPCA, a calendar-structured sparse principal component method for interpretable representation learning in multi-periodic electricity consumption profiles, demonstrating high explained variance with increased sparsity and coherence.
DynG-Diff proposes a state-aware dynamic guidance diffusion framework for probabilistic multivariate time series forecasting, improving robustness by adaptively handling variable heterogeneity.
This paper critically audits benchmark datasets and baselines for spatiotemporal forecasting, showing that classical linear models often compete with graph neural networks, highlighting issues with discriminative reliability and advocating for more rigorous evaluation.
The paper proposes SCPaT, a Transformer-based framework for multivariate time series forecasting that uses semantic structured partitioning to address limitations in existing patch-based methods and demonstrates effectiveness on real-world datasets.
This paper introduces DNBNet, a debiased neural basis-function network for irregular time series forecasting, addressing limitations in existing methods by correcting asymptotic bias and using adaptive neural basis functions.
This study uses connected vehicle telemetry data from Sydney to predict high-risk driving hotspots, benchmarking machine learning and time-series models for proactive road safety interventions.
A comprehensive guide explaining the Kalman filter and its application in building smarter trading systems, including mathematical foundations and production-grade examples.
This academic paper presents a hierarchical ensemble pipeline for anomaly detection in ESA satellite telemetry, utilizing shapelet-based and statistical feature extraction to identify subtle anomalies in multivariate time-series data.