ML Resources

Easily the best way to improve at machine learning is to just get practice at it. Practice helps you improve at machine learning in two ways: 1. You get familiar with implementing common algorithms. Whether you use machine learning libraries or roll your own, becoming comfortable with these algorithms is critical. 2. By working with a variety of data sets, you can also gain an intuition of how data is structured, so you know when to apply what algorithms.

So, the most effective way to gain experience at machine learning is to experience implementing a variety of algorithms over differently behaved data sets.

I've consolidated the machine learning resources I am familiar with, both from school and outside, into a list below. I've marked each data set with what model(s) I used for the data set, but please experiment with any algorithm that interests you!



Note: For any supervised classification problem, you can also use it to practice unsupervised clustering by ignoring the class labels.

Unsupervised Learning

Dimensionality Reduction