The rigorous, mathematical machine-learning course that many practitioners consider the real thing. CS229 goes deep into the theory behind the algorithms, with serious maths. The lecture videos and notes are free, but it is demanding and there is no certificate from the free version.
Good for: Learners with strong maths who want the theory behind machine learning, not just how to call a library.
Less suitable if: You are new to machine learning, or you want an applied, code-first course.
You are comfortable with calculus, linear algebra and probability
You want the maths behind ML, not just the tools
You can study demanding material independently
Consider something else if you…
You are new to machine learning
You want an applied, beginner-friendly course
You need a certificate
Requirements: Solid Python, multivariable calculus, linear algebra and probability. This is advanced.
Realistic time: About 55 hours of lectures, plus substantial problem sets.
About this course
CS229 is Andrew Ng's graduate-level Stanford ML course — the academic original that preceded his Coursera courses and goes significantly deeper into the mathematical foundations. Where the Coursera ML Specialization builds intuition and implements in Python, CS229 works through full probabilistic derivations and proofs: maximum likelihood estimation for supervised learning, EM algorithm for unsupervised learning, support vector machines and kernels, principal component analysis, reinforcement learning with MDPs and Q-learning, and the theoretical foundations of learning guarantees.
What you'll learn
Derive supervised learning algorithms from probabilistic first principles
Understand and apply the EM algorithm for unsupervised learning
Work with support vector machines and kernel methods mathematically
Apply reinforcement learning with MDPs and Q-learning
Understand PAC learning and generalization bounds
This course includes
55h
On-demand video
Yes
Mobile access
English
Language
What it costs
The lecture videos and course notes are free — the older recorded version is on Stanford Engineering Everywhere, and current notes are public. Stanford's paid online version costs several thousand dollars, but you do not need it to learn.
Comparison · LBS
Compare alternatives for CS229: Machine Learning
Same topic, different options. We surface the trade-offs others hide so you can pick the course that actually fits your time, budget, and goals.
SOStanford Online4.9(11,000)
CS229: Machine Learning
Price
Free
Free lecture materials available online
Duration
55 hrs
Level
Advanced
Certificate
COCoursera4.8(14,000)
Google Advanced Data Analytics Professional Certificate
The lecture videos and notes are free to access. Stanford's paid online enrolment costs several thousand dollars, but the learning materials themselves are open.
CS229 is the rigorous, maths-heavy Stanford course. The Coursera Machine Learning Specialization is a gentler, applied version of the same ideas, aimed at beginners.
Comfort with Python, multivariable calculus, linear algebra and probability. It is genuinely advanced.
Not from the free materials. Only Stanford's paid enrolment leads to a formal record.
Yes if you want the theory and have the maths. If you want to get building quickly, start with an applied course instead.
Yes. It is one of the more demanding ML courses, largely because of the maths.