machine-learning-course-by-andrew-ng-in-rust

Week 1

week1 dir contains linear regression demo for the first week.

Just run it, click to create additional points and see how it converges.

Week 2

week2 dir contains linear regression with multiple features demo for the second week. The implementation works with any number of features, but we use 2 because human beings have troubles perceiving more than 3 dimensions.

You can left click the image and rotate it with the mouse and add points by the inputs on the bottom.

There 2 algorithms: linear regression and normal equation and you can switch between them by changing USE_NORMAL_EQUATION in src/constants.rs file

Week 3

week3 dir contains a demo for logistic regression with 1 and 2 features. For 2 features there's a 2d visualization where x, y and size of the dot are original data. For 3 features we visualize points in 3d space and prediction function as a sigmoid surface. You can add new points and see how the surface will get adjusted.

Week 4

There's no demo for week 4 since week 5's demo includes the material from week 4.