Free Training Resources for AI and Machine Learning

Training course
Credit: Shutterstock

AI and machine learning are splashed across the headlines these days, for good reason. AI has filtered into everyday life. Uber and Lyft use AI to determine the price of a ride, as do autopilots on commercial airplanes. Even spam filters use a form of AI.

Big high-tech firms have been pursuing better AI technologies for years — IBM Watson, Microsoft Project Oxford and Google DeepMind are just a few examples. Analyst firm IDC estimates that, by 2018, more than 70 percent of developer teams will include cognitive/AI functionality in one or more applications. All of this points to a wealth of job opportunities in AI and machine learning, and the need for training and skills development to arm IT professionals to fill such job slots.

These 10 sources of free AI and machine learning training can get you started on a new career path. Are your algebra, calculus and probability skills up to speed? You'll need them for some of the academic courses listed here, and routinely include in AI and machine learning curriculum.

Stanford University

Stanford University

Stanford's CS229 - Machine Learning course, offered as part of the Stanford Engineering Everywhere program, dives into supervised and unsupervised learning, learning theory, reinforcement learning and adaptive control. You can watch all 20 lectures online or download them, and then try your hand at the assignments (solutions are provided so you can check your work).

The CS221: Artificial Intelligence: Principles and Techniques course, taught by Percy Liang, is not actually freely available but some of its course materials are. Although you need a Stanford ID and password to access the online lecture videos, scroll down the course page to the Schedule section for links to PowerPoint slides that contain the instructor's notes.

UC Berkeley

UC Berkeley

University of California Berkeley's simple CS188 Intro to AI web page provides links to 25 lecture presentations (hosted on YouTube) as well as all lecture slides. If you want to work through the online homework assignments you must sign up with edX. But anyone can take part in the Pac-Man projects, in which you play games, run into issues and then modify code to solve them. The purpose of the Pac-Man projects is to help you learn AI concepts like informed state-space search, probabilistic inference and reinforcement learning.

Another course to check out is CS 294: Deep Reinforcement Learning. In addition to online lectures and slides, there is a ton of free materials to browse through.

MIT

MIT

MIT's free Artificial Intelligence course, taught by Patrick Henry Winston, is from 2010 but still popular with many AI beginners. More than 20 lecture video presentations cover basic AI concepts, problem solving and learning methods. The course includes interactive demonstrations designed to "help students gain intuition about how artificial intelligence methods work under a variety of circumstances." You can also watch the lecture videos on Apple iTunes.

Fast.ai

Fast.ai

Jeremy Howard, founder of Enlitic, teaches a certificate course at The Data Institute at USF, which is also available as a free, seven-week Practical Deep Learning for Coders online course through fast.ai. If you've got at least one year of coding under your belt, and can commit to about 10 hours of class work per week, you're a good candidate for this course. You'll learn how to create computer vision models, and explore natural language processing and recommendation systems.

Udacity

Udacity

Udacity offers several AI, machine learning and deep learning courses. Consider starting with Intro to Artificial Intelligence (four-month course) and then Intro to Machine Learning (10-week course), both of which are considered intermediate level. Seasoned engineers and data scientists should check out Deep Learning by Google, an advanced-level course on designing intelligent systems that learn from large, complex datasets. The course was created in part by Vincent Vanhoucke, principal scientist at Google and the Google Brain team technical lead. You might also be interested in Artificial Intelligence for Robotics by the Georgia Institute of Technology.

edX

edX

The Artificial Intelligence (AI) by ColumbiaX starts with an intro to AI and its history. Then it covers how to build intelligent agents, and explores machine learning algorithms, natural language processing, robotics and vision. Students also learn how to solve AI problems using Python. The course is free, but students must pay $300 to get a verified certificate upon completion.

Coursera

Coursera

The Artificial Intelligence (AI) by ColumbiaX starts with an intro to AI and its history. Then it covers how to build intelligent agents, and explores machine learning algorithms, natural language processing, robotics and vision. Students also learn how to solve AI problems using Python. The course is free, but students must pay $300 to get a verified certificate upon completion.

Nvidia

Nvidia

Known for its graphics processors and visual computing technologies, Nvidia started its Deep Learning Institute (DLI) to give developers, data scientists and researchers hands-on experience solving problems with deep learning. In 2017, the company launched an initiative to up-skill 100,000 developers in deep learning by 2018, and released several free and low-cost ($30) labs to support the effort. Browse all available labs on the company's Online Self-Paced Labs page.

Kaggle Competitions

Kaggle Competitions

Kaggle is a platform for predictive modeling. One of the cool things it offers is analytics competitions that challenge researchers and data miners to create models for predicting and describing data. Organizations sponsor challenges and provide prize money in some cases, to the tune of $15,000 to more than $1 million. Recent challenges included the Passenger Screening Algorithm Challenge for the Department of Homeland Security, Zillow Prize: Zillow's Home Value Prediction (Zestimate), and Planet: Understanding the Amazon from Space. Beginners should start by taking the Titanic: Machine Learning from Disaster tutorial.

OpenAI Gym

OpenAI Gym

Tesla and SpaceX leader Elon Musk's OpenAI Gym (in beta as of this writing) aims to provide a playground of sorts for "developing and comparing reinforcement learning algorithms." Using Python and frameworks like TensorFlow or Theano, gym-goers write algorithms, then share results with the community for review and feedback. OpenAI Gym provides a tutorial to get started, and offers learning experiences that involve playing Atari games and board games, and controlling a simulated robot.

YouTube

YouTube

Most of us use YouTube as a go-to source for free training, and there's plenty of AI and machine learning tutorials and courses waiting to be consumed. But I have a couple of favorites. Artificial Intelligence: Machine Learning Introduction is a great general introduction to AI, intelligent systems and machine learning algorithms. Machine Learning Recipes is a series of high-quality video tutorials by Josh Gordon, developer advocate for Google. Don't just browse our picks – you can easily find a plethora of videos by searching for artificial intelligence training or machine learning training on the main YouTube page.