Introduction to Machine Learning

Course Overview

Discover how computers learn from experience in this 25-lecture course by Professor Michael L. Littman. Master Python-based machine learning applications from recommender systems to deep neural networks, while exploring ethical implications of AI.

Video Lessons

Foundations

1. Telling the Computer What We Want (31 min)
   Overview of machine learning history and medical diagnosis applications
2. Starting with Python Notebooks and Colab (17 min)
   Setup guide for browser-based Python programming

Core Algorithms

3. Decision Trees for Logical Rules (31 min)
   Building predictive models for spelling patterns and diabetes risk
4. Neural Networks for Perceptual Rules (30 min)
   Image/audio processing comparisons between decision trees and neural nets
5. Opening the Black Box of a Neural Network (29 min)
   Python implementation of neural networks for green screen effects

Statistical Approaches

6. Bayesian Models for Probability Prediction (29 min)
   Naïve Bayes applications in spam filtering
7. Genetic Algorithms for Evolved Rules (28 min)
   Evolutionary approaches to optimization problems
8. Nearest Neighbors for Using Similarity (29 min)
   Comfort prediction and malware detection systems

Practical Challenges

9. The Fundamental Pitfall of Overfitting (28 min)
   Avoiding model bias in diabetes dataset analysis
10. Pitfalls in Applying Machine Learning (28 min)
    Case studies in medical and legal data misinterpretation

Advanced Techniques

11. Clustering and Semi-Supervised Learning (27 min)
    Combining labeled/unlabeled data strategies
12. Recommendations with Three Types of Learning (30 min)
    Conference paper review and Netflix prize case studies
13. Games with Reinforcement Learning (30 min)
    AI strategies for Othello, chess, and Go

Deep Learning

14. Deep Learning for Computer Vision (27 min)
    ImageNet challenge and object recognition systems
15. Getting a Deep Learner Back on Track (30 min)
    Debugging animal classification neural networks

Language Processing

16. Text Categorization with Words as Vectors (30 min)
    Word embeddings and spelling correction systems
17. Deep Networks That Output Language (29 min)
    Machine translation and story generation

Creative Applications

18. Making Stylistic Images with Deep Networks (29 min)
    Dual-stage creative processes in image generation
19. Making Photorealistic Images with GANs (30 min)
    Generative adversarial networks for image manipulation

Emerging Frontiers

20. Deep Learning for Speech Recognition (30 min)
    Evolution of voice interaction systems
21. Inverse Reinforcement Learning from People (29 min)
    Behavioral prediction algorithms
22. Causal Inference Comes to Machine Learning (30 min)
    Moving beyond correlation to causation

Future Directions

23. The Unexpected Power of Over-Parameterization (30 min)
    2015 deep-learning breakthroughs
24. Protecting Privacy within Machine Learning (31 min)
    Data anonymization challenges and solutions
25. Mastering the Machine Learning Process (34 min)
    Meta-learning and the future of AI