A practical, top-down introduction to deep learning taught by fast.ai, built around writing code and training models early before the underlying theory. It covers applications such as computer vision, NLP, and tabular data using a high-level library.

Commenters recommend it as the place to start, call it beginner-friendly and almost paint-by-numbers, and suggest doing projects from it before going back to the math.

On Reddit (2)

1. Recommendsmano-vijnanar/learnmachinelearning2022in “Help with getting started ML”

   Start with FastAI's Deep Learning for Coders, do some projects, and only then go back and learn math (Coding the Matrix or MIT's open courseware course for linear algebra, and MIT's calculus courses if necessary. Otherwise, if you're getting by fine with a more shallow understanding just watch…

2. RecommendsToothpasteTimebombr/learnmachinelearning2019in “Good projects for beginners”▲ 4

   Check out fast.ai. They’ve got a whole beginner-oriented course on computer vision. It’s crazy how simple they have made this stuff. Almost paint-by-numbers easy.

Related courses

• 18.06 Linear Algebra (Gilbert Strang)MIT OpenCourseWare · 78 recommenders
• Coding the Matrix: Linear Algebra through Computer Science ApplicationsCoursera · 2 recommenders
• Machine Learning Specialization (Andrew Ng Deep Learning Specialization)Coursera · 3 recommenders
• Machine Learning (Andrew Ng)Coursera · 240 recommenders
• Practical Deep Learning for Codersfast.ai · 112 recommenders
• CS231n: Convolutional Neural Networks for Visual RecognitionStanford University · 39 recommenders
• Deep Learning Specialization (Andrew Ng)Coursera · 31 recommenders
• Neural Networks for Machine Learning (Geoffrey Hinton)Coursera · 25 recommenders