Skip to content

I use a self-implemented Trust-Region-Method to solve the optimization problem and calculate the accuracy based on test data

Notifications You must be signed in to change notification settings

Paulnkk/Columbia-IEOR-4004

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 

Repository files navigation

Solving-L2-Logistic-Regression-Problem

In the course "Optimization Models and Methods" I implemented a trust-region method which solves the L2 regularized Logistic Regression Problem. The Dogleg method is used to solve the subproblems. The target parameter, beta^(star), is calculated using a Maximum Likelihood approach.

I have implemented the objective function as well as its gradient and Hessian matrix. In addition, I used the following parameters to initialize the method: lambda = 10^(-5), beta^0 = 0 (starting point), t^0 = 1 (starting radius), t^(hat) = 10^4 (maximum radius), Eta = 0.05 and termination tolerance = 10^(-5).

About

I use a self-implemented Trust-Region-Method to solve the optimization problem and calculate the accuracy based on test data

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages