Learning objectives

• Locate the online resources for learning machine learning.

Introduction

There are numerous resources online to learn machine learning by yourself, and many are free. I will summarize a few from different aspects.

Online courses

• Stanford Engineering Everywhere offers the free videos, lecture notes, homework problems and solutions, of a collection of most popular computer science and engineering classes at Stanford University. CS 229 Machine Learning taught by Andrew Ng is the a must-have for every machine learning beginner.

• Coursera is among the best websites offering high-quality online classes. There are a series of machine learning classes offered by Stanford University or deeplearning.ai that are highly recommended by numerous students and researchers. With a free account on Coursera, one already has full access to the course videos and quiz. These classes are offered at different levels and meet the need for both beginners and seasoned machine learning researchers.

Machine learning packages

Keras, Tensorflow, scikit-learn are open source python packages for machine learning. There are many tutorials and examples online for these packages making self-learning easier than ever.

Books and reading materials

• The first material I recommend is the lecture notes of Stanford CS 299. It is concise, modern, and light-weighted compared to classical statistical learning books, and is perfect for beginners.

• If you are more interested understanding machine learning from the perspective of statistics, there are some classical books, such as The Elements of Statistical Learning by Trevor Hastie, Robert Tibshirani, and Jerome Friedman. This book is officially released as PDF format and free.

• For Chemistry researchers, there is a new book by Jon Paul Janet and Heather Kulik focused on Machine learning in chemistry. This book is available from ACS.

Data Analysis

• Practical machine learning research involves a lot of data analysis. There is a series of data analysis tutorials offered by the Molecular Science Institute (MolSSI) focused on python and pandas.