Home/Programming/Python/Python for Everybody

Machine Learning Specialization

4.9(78,000)·720K enrolled·Updated February 2025
Intermediate 94 hours English Professional CertificateFREE
Editor's Pick
The single best ML foundation course. Modernised in 2022 with Python (not Octave). If you do one ML course, do this one.

What you'll learn

Build ML models in Python using NumPy and scikit-learn
Train supervised models: linear & logistic regression, neural networks
Apply unsupervised learning: clustering, anomaly detection, recommenders
Build a deep learning model with TensorFlow
Diagnose bias and variance and apply ML best practices
Use decision trees, random forests and gradient boosting
Build a recommender system with collaborative filtering
Apply reinforcement learning to a lunar lander problem

This course includes

94h
On-demand video
100+
Readings
30+
Programming labs
30
Graded quizzes
Yes
Certificate
Comparison · LBS

Compare alternatives for Python

Same topic, different providers. We surface the trade-offs others hide so you can pick the course that actually fits your time, budget, and goals.
Coursera4.9(78,000)
Machine Learning Specialization
Price
FREE
Audit free · Cert $49/mo
Duration
94 hrs
Level
Intermediate
Certificate
Professional
Coursera4.6(102K)
IBM Data Science Professional Certificate
Price
FREE
Audit free · Cert $49/mo
Duration
110 hrs
Level
Beginner
Certificate
Professional
DataCamp4.7(14,800)
Data Scientist with Python Career Track
Price
$29/mo
$29/mo
Duration
88 hrs
Level
Beginner
Certificate
Completion
Prices and ratings refreshed daily. We're not affiliated with any single provider.

Syllabus· 3 courses · 15+ lessons

Expand all →
  • Linear regression with one variableVideo · 45 min
  • Gradient descent intuitionVideo · 35 min
  • Multiple linear regressionLab · 90 min
  • Logistic regression for classificationVideo · 40 min
  • Regularization to address overfittingLab · 75 min
  • Neural networks: intuition and implementationVideo · 50 min
  • TensorFlow implementationLab · 90 min
  • Vectorisation and matrix mathReading · 30 min
  • Decision trees and ensemblesVideo · 45 min
  • XGBoost in practiceLab · 75 min
  • K-means clusteringVideo · 35 min
  • Anomaly detectionLab · 60 min
  • Collaborative filtering recommendersVideo · 40 min
  • Content-based filtering with neural netsLab · 75 min
  • Reinforcement learning lunar landerProject · 3 hours

Instructor

AN
Andrew Ng
Founder, DeepLearning.AI · Adjunct Professor, Stanford
8.0M learners22 courses 4.9 instructor rating

Andrew Ng is one of the most influential AI educators in the world. His original "Machine Learning" course on Coursera has over 4.8 million enrolments. This three-course Specialization is the 2022 modern remake co-produced with Stanford and DeepLearning.AI.

Requirements

  • Basic Python (lists, loops, functions) — can be learned alongside
  • High-school maths: algebra, basic derivatives
  • A computer that can run Jupyter notebooks
  • 8–10 hours per week for about 3 months

Who this course is for

  • Software engineers moving into ML
  • Data analysts upskilling to data scientist
  • STEM students wanting an ML foundation
  • Anyone who took the original 2012 ML course and wants the modern remake

About this provider

C
Coursera
University-backed online learning platform · 142M learners · 7,000+ courses
4.6 trust score·Refund within 14 days
Browse all Coursera courses →

Frequently asked questions

This Specialization is the modern replacement — same teacher, but Python instead of Octave, modern libraries (TensorFlow, scikit-learn), and updated coverage of deep learning, XGBoost and reinforcement learning. The 2012 course is being deprecated.
Take this one first. It covers the ML foundations (regression, classification, neural nets). Deep Learning Specialization assumes you know this material and goes deeper into CNNs, RNNs and Transformers.
No. Andrew explicitly designed this for engineers — calculus and linear algebra are introduced as needed. If you can read basic math notation you will be fine.
Yes if you are job hunting in ML — Andrew Ng + Stanford + DeepLearning.AI is one of the strongest signals on a resume. Most learners finish in 2–3 months, so total cost is $100–$150.
Audit it for free — you get all videos and readings. You only lose graded labs. Combine with fast.ai's free Practical Deep Learning course for project work.
Free
Cert $49/mo
View on Coursera