Y2HL OrbitYes to Happy Learning

Courses / ML

Python · Undergraduate, no prerequisites

Machine Learning Fundamentals

by Dr. Asesh Kumar Tripathy

Decision trees, Bayesian learning, neural networks and beyond, with programs that run in your browser. 35 topics from decision trees to Q-learning, each with a clear explanation, Python programs you can run and edit on the page, practice problems and a self-marking quiz.

  • 35topics
  • 280quiz questions
  • 5CO tests
  • 39programs that run on the page
Start Topic 1 How to study this course

The 35 topics

CO1Machine learning basics and decision trees

Topic 1 What machine learning is: well-posed learning problems Topic 2 Designing a learning system Topic 3 Decision trees: entropy and information gain Topic 4 The ID3 algorithm Topic 5 Hypothesis space search and inductive bias Topic 6 Issues in decision tree learning CO1 test 20 questions from these topics

CO2Bayesian learning

Topic 7 Bayes theorem and its notation Topic 8 Bayes and concept learning: MAP and ML hypotheses Topic 9 Maximum likelihood, least squares and predicting probabilities Topic 10 MDL, the Bayes optimal classifier and the Gibbs algorithm Topic 11 Naive Bayes classifier and learning to classify text Topic 12 Bayesian belief networks Topic 13 The EM algorithm and k-means CO2 test 20 questions from these topics

CO3Neural networks and genetic algorithms

Topic 14 Neural network representations and the perceptron Topic 15 Perceptron training, gradient descent and the delta rule Topic 16 Multilayer networks Topic 17 The backpropagation algorithm Topic 18 Genetic algorithms Topic 19 Genetic programming and models of evolution CO3 test 20 questions from these topics

CO4Rule learning, analytical learning and reinforcement learning

Topic 20 Learning sets of rules: sequential covering Topic 21 Learning first-order rules: FOIL Topic 22 Induction as inverted deduction and inverting resolution Topic 23 Analytical learning: inductive and analytical problems Topic 24 Learning with perfect domain theories: Prolog-EBG Topic 25 Reinforcement learning: Q-learning CO4 test 20 questions from these topics

CO5Machine learning programs in Python

Topic 26 Data exploration and visualisation with pandas Topic 27 Data preprocessing and handling missing values Topic 28 Linear regression Topic 29 Logistic regression with gradient descent Topic 30 Decision trees in practice: evaluation and post-pruning Topic 31 Random forests Topic 32 Naive Bayes on census-style income data Topic 33 K-nearest neighbours Topic 34 K-means clustering in practice Topic 35 Building an ANN classifier, with a genetic optimiser CO5 test 20 questions from these topics

See my marks →

How to study this courseThe study rhythm for each topic, how the programs on the page work, how the quizzes and CO tests work, and how your marks are kept. Lab and skill programsWhich topic covers each lab session and skill session, and what to build for each. Project guideHow to take a machine learning project from a problem statement to a demonstration, one session at a time, plus how to use TensorFlow on your own computer.

Course overview

Machines learn from data. In this course you find out how, and build the programs yourself.

This course takes you through the main ways a computer can learn from experience:

Every topic page explains the idea in plain words, works through an example, and gives Python programs that run right on the page. Press Run, change something, and run it again. You don't need to install anything to start.

Course outcomes

CO You will be able to Level
CO1 Apply machine learning techniques using decision trees to solve real-world problems. Apply
CO2 Build Bayesian models for solving classification and prediction problems. Analyse
CO3 Apply neural network and genetic algorithm techniques to solve classification and prediction problems. Apply
CO4 Demonstrate learning first-order rules, analytical learning, explanation-based learning and reinforcement learning. Understand
CO5 Implement machine learning techniques using the Python language. Evaluate

Syllabus

Textbooks

Topic pages refer to these books by number.

  1. Textbook 1: Machine Learning, McGraw Hill, 1997.
  2. Textbook 2: Machine Learning: An Algorithmic Perspective, CRC Press, 2009.
  3. Textbook 3: Core Python Programming, 2nd edition, Pearson, 2007 reprint.

Reference books:

  1. Introduction to Machine Learning, The MIT Press, 2010.
  2. Programming Python, O'Reilly.

Free online courses