Getting Started with Machine Learning— Fall ‘17
Over on LinkedIn, there is a discussion on the New York Times article, which states that entry level PhD’s or people with a few years of experience with machine learning are making millions from a shortage of AI skilled labor. Dozens of people are asking for good resources to get started. This is by no means a compendium, but by the end of it, you should have plenty of experience to find your own. These are just my picks that have helped me get moving. Think of it as your first semester before a lifetime of learning.
Online Courses
Geoffrey Hinton’s course on neural networks. He basically started this whole field. Fun fact from Wikipedia: He is the great-great grandson of George Boole!
This course on Udemy walks you through several types of neural networks and more using neat tools, like Keras, with homework after each section to help you understand the structure and flow of data. It’s paced very well. I feel like courses move too slowly, but this struck the right balance.
This is a snapshot of the course from 2006, so don’t expect to use the latest tools, but this is a great course to go deeper into what modern tools do today. This will give you a deeper grasp on models, classification, and the math that powers ML. Emphasis on the math.
Harrison, of recent Python plays GTA5 fame, has hours of in-depth Python tutorials that follow a terrific pattern of building a basic version from scratch, and using the popular tool sets. Honestly, it was this story hitting Reddit earlier this year that got the gears turning in my head about machine learning.
Web
More of a collection of ‘getting started’ pages and tools than ‘tutorials or courses’. These are great to get up to speed, and if you are constantly referring back to them, your probably doing things right. Don’t stress.
Book
Ok, so there’s just one that I’ve read so far about ML (web resources are just so rich):
It’s a brief book, but I appreciate how it covers the mathematical interplay between layers of a neural net. It helped me see neural nets in a very clear way and spawned a couple of projects I’ve been working on. It’s also $3.98 on Kindle. Neural network knowledge for less than a latte.
https://www.amazon.com/Algorithms-Live-Computer-Science-Decisions/dp/1627790365
Algorithms to Live By might help you spark an idea about applying ML to your life or generate questions to try to answer. This was one of my favorite recent reads.
That should get you started
I’m not an expert in machine learning (yet). I’m just really passionate about technology, and this is the new, new thing. Take the time in the middle of all of these to pursue an idea that pops up. If you’re like me, the real learning comes through trial and error. Try to encode eBay listings for classification to predict the final auction price. Try to find the average value of a home in your city based on location. These were random thoughts I had while I worked through these resources. Try to ask interesting questions and fit them into data.
There are an ocean of things to learn, and if you have any recommendations, please comment below. Thanks for reading.
