Resources for getting into machine learning and statistics

As the inaugural blog post here, I thought I could share some of the resources that helped me when I first started getting into ML and stats.

• https://machinelearningmastery.com/ This blog is developed by a guy who creates code snippets and explains important machine learning concepts in a very easy to understand manner.

• https://www.udacity.com/course/machine-learning–ud262 This course was among the first I took which explained much of the fundamentals about machine learning. As a support for the course, and as a parallel resource, the lecturer for my class (Formerly prof, now dean) Dr. Charles Isbelle created an online course with essentially the same content.

• https://www.deeplearningbook.org/ Ian Goodfellow’s book is criminally underrated, but a good resource for going into the really deep learning aspects of machine learning. It’s probably the case that “shallow” learners are probably more suited for 95% of use cases, but this book is still pretty useful for essentially summarizing decades of research and some of the unspoken art of the craft into something that is good for a beginner/intermediate ML guy like me.

• http://www.feat.engineering/ Feature engineering and data cleaning are entirely skipped over by some courses or given way too little time in other course. In my experience, understanding the features and the data can take up to 90% of the time in a project if not more. This book is a good primer on a few common feature engineering techniques.

• Applied Statistics and Probability for Engineers This book wasn’t really useful for ML perse, but it was a good primer on introductory principles in statistics. It really helps to be a primer on later concepts in statistics.

Of note, while linear algebra is a major part of some of the more advanced concepts in ML, I haven’t really had to use any of it to really implement algorithms on my own that often. As a result, I’m quite fine to understand the basic concepts but to skip over the formal maths for now.