Multiplicative Algorithm for Orthgonal Groups and Independent Component Analysis
Toshinao Akuzawa
Abstract
The multiplicative Newton-like method developed by the author et al. is extended to the situation where the dynamics is restricted to the orthogonal group. A general framework is constructed without specifying the cost function. Though the restriction to the orthogonal groups makes the problem somewhat complicated, an explicit expression for the amount of individual jumps is obtained. This algorithm is exactly second-order-convergent. The global instability inherent in the Newton method is remedied by a Levenberg-Marquardt-type variation. The method thus constructed can readily be applied to the independent component analysis. Its remarkable performance is illustrated by a numerical simulation.
Create a lesson
Related papers
AI Models Can Provably Hide Arbitrary Capabilities
Andis Draguns, Sumeet Ramesh Motwani, Raymond Douglas et al.
KL Before and After Mixing: A Round-Complexity Separation in Parallel Sampling
Tianyi Wu
Designing Protein Structures from Natural Language
Joseph Lee, Yanjing Li, Bo Qiang et al.
Revisiting Tabular Deep Regression: Objectives and the Limits of Global Early Stopping
Danil Rudenko, Georgii Bulgakov, Nikolay Kartashev et al.
Prediction Is Not a Sufficient Memory State
Robert Jenkinson Alvarez
Partially Informed Stochastic Optimization via Zeroth-Order Residual Tracking
El Mahdi Chayti, Alireza Mirrokni, Martin Jaggi