Adaptive Iterative Soft-Thresholding Algorithm with the Median Absolute Deviation
By: Yining Feng, Ivan Selesnick
Potential Business Impact:
Makes computer math problems solve themselves.
The adaptive Iterative Soft-Thresholding Algorithm (ISTA) has been a popular algorithm for finding a desirable solution to the LASSO problem without explicitly tuning the regularization parameter $\lambda$. Despite that the adaptive ISTA is a successful practical algorithm, few theoretical results exist. In this paper, we present the theoretical analysis on the adaptive ISTA with the thresholding strategy of estimating noise level by median absolute deviation. We show properties of the fixed points of the algorithm, including scale equivariance, non-uniqueness, and local stability, prove the local linear convergence guarantee, and show its global convergence behavior.
Similar Papers
Edge-preserving Image Denoising via Multi-scale Adaptive Statistical Independence Testing
CV and Pattern Recognition
Cleans up blurry pictures, finding sharp edges.
An Improved Optimal Proximal Gradient Algorithm for Non-Blind Image Deblurring
CV and Pattern Recognition
Cleans up blurry pictures better than before.
Fast sparse optimization via adaptive shrinkage
Optimization and Control
Makes computer math problems solve much faster.