cvxopt

Convex Optimization: Assignment and Project
Runtian Zhai (ID: 1600012737)

What is this?

This repository contains codes for assignments of the course Convex Optimization by Professor Wen. You can view the reports online at http://www.runtianz.cn/cvx

How to run the codes?

For a quick start, simply run Test_l1_regularized_problems.m in the root folder and all methods will be automatically tested. example_output.txt contains an example output of the program.
A list of methods follows:

| Filename | Method | | ------ | ------ | | l1_pgd.m | Projected gradient method | | l1_subgradient.m | Subgradient method | | l1_smooth.m | Gradient method for smoothed primal | | l1_smoothd.m | Smoothing with decreasing lambda | | l1_smoothf.m | Smoothing with fast gradient method | | l1_smooth(1,2).m | Other smoothers | | l1_proximal.m | Proximal gradient method | | l1_fistab.m | FISTA: Basic version | | l1_fistad.m | FISTA: Descent version | | l1_nes2.m | Nesterov's 2nd method | | l1_admm.m | ADMM for dual | | l1_admmlin.m | ADMM with linearization for primal | | l1_momentum.m | Subgradient with momentum | | l1_adagrad.m | AdaGrad | | l1_rmsprop.m | RMSProp | | l1_adam.m | Adam | | l1_cvx_mosek.m | Calling mosek from cvx | | l1_mosek.m | Calling mosek directly | | l1_cvx_gurobi.m | Calling gurobi from cvx | | l1_gurobi.m | Calling gurobi directly |

Acknowledgement

I would like to thank Professor Wen for this excellent course. I would also like to thank TAs Jiang and Haoming, who spend their time reading these assignments.