Independent Component Analysis
Independent Component Analysis (ICA) is a statistical method used to separate a multivariate signal into additive subcomponents that are statistically independent, revealing underlying sources from their mixed observations. Current research focuses on extending ICA's capabilities to handle nonlinear relationships, high-dimensional data, and non-Gaussian sources, often employing neural networks and incorporating concepts like sparsity and structural constraints to improve identifiability and interpretability. These advancements are significantly impacting fields like signal processing, time series forecasting, and neuroscience by enabling more accurate feature extraction, improved model performance, and deeper insights into complex systems.
Papers
Understanding Higher-Order Correlations Among Semantic Components in Embeddings
Momose Oyama, Hiroaki Yamagiwa, Hidetoshi Shimodaira
TSI: A Multi-View Representation Learning Approach for Time Series Forecasting
Wentao Gao, Ziqi Xu, Jiuyong Li, Lin Liu, Jixue Liu, Thuc Duy Le, Debo Cheng, Yanchang Zhao, Yun Chen