The field of time series analysis and forecasting is rapidly evolving, with a focus on developing innovative methods to handle complex and irregular data. Recent research has emphasized the importance of effective time series imputation, anomaly detection, and probabilistic forecasting. Novel approaches, such as implicit neural representations and Fourier adaptive noise-separated diffusion, have shown promise in improving forecasting accuracy and reducing uncertainty. Additionally, hybrid models combining traditional statistical methods with machine learning techniques have demonstrated superior performance in certain applications. Noteworthy papers in this area include ImputeINR, which proposes a novel approach to time series imputation using implicit neural representations, and FALDA, which introduces a probabilistic framework for time series forecasting using Fourier adaptive noise-separated diffusion. Overall, the field is moving towards more sophisticated and robust methods for analyzing and forecasting complex time series data.
Advances in Time Series Analysis and Forecasting
Sources
ImputeINR: Time Series Imputation via Implicit Neural Representations for Disease Diagnosis with Missing Data
Inferring the Most Similar Variable-length Subsequences between Multidimensional Time Series
Anomaly Detection for Non-stationary Time Series using Recurrent Wavelet Probabilistic Neural Network
IISE PG&E Energy Analytics Challenge 2025: Hourly-Binned Regression Models Beat Transformers in Load Forecasting
Driving Mechanisms and Forecasting of China's Pet Population-An ARIMA-RF-HW Hybrid Approach
Effective Probabilistic Time Series Forecasting with Fourier Adaptive Noise-Separated Diffusion
Rethinking Irregular Time Series Forecasting: A Simple yet Effective Baseline
Deep Learning-Based Forecasting of Boarding Patient Counts to Address ED Overcrowding
Continuous Representation Methods, Theories, and Applications: An Overview and Perspectives