Blockwise Coordinate Descent Procedures For The Multi-Task Lasso, With Applications To Neural Semantic Basis Discovery

Free registration required

Executive Summary

This paper develops a cyclical blockwise coordinate descent algorithm for the multi-task Lasso that efficiently solves problems with thousands of features and tasks. The main result shows that a closed-form Winsorization operator can be obtained for the sup-norm penalized least squares regression. This allows the algorithm to find solutions to very largescale problems far more effciently than existing methods. The author demonstrates how this learned basis can yield insights into how the brain represents the meanings of words.

  • Format: PDF
  • Size: 251.4 KB