Unit 8 Overview: Machine Learning Foundations

Unit 8: Machine Learning Foundations (Capstone) — Overview

Connection to Prior Units: This is the capstone unit of CSC 114. Over the past seven weeks you have built a complete toolkit for artificial intelligence: you formulated problems as search (Unit 3), optimized solutions under constraints (Unit 4), reasoned with formal logic (Unit 5), built knowledge-based inference systems (Unit 6), and handled uncertainty with probability (Unit 7). Machine learning draws on every one of those ideas. Before starting Unit 8, make sure you are comfortable with Bayesian reasoning (7.4) and the notion of probability as a model of uncertainty — ML is fundamentally probabilistic.

What This Unit Is About

Every AI system you have studied so far required a human expert to encode knowledge — writing search heuristics, specifying logic rules, or designing probability tables. Machine learning asks a different question: what if the system could discover those rules on its own by studying examples?

Tom Mitchell’s 1997 definition remains the standard: "A computer program is said to learn from experience E with respect to some task T and performance measure P, if its performance on T, as measured by P, improves with experience E." In plain language — the system gets better at its job the more data it sees. This week you will study the foundations of that idea: how machines learn from labeled examples (supervised learning), how we build and read decision trees, how we measure whether a model actually generalizes, and how machine learning fits into the broader landscape of modern AI.

Because this is the final unit of the course, the week also serves as a synthesis: you will see how search, optimization, logic, knowledge systems, and probability all contributed to making modern machine learning possible — and where the field is going next.

Learning Objectives

By the end of this unit, you will be able to:

  1. Define machine learning and distinguish it from traditional programming

  2. Explain the three paradigms of ML: supervised, unsupervised, and reinforcement learning

  3. Describe the supervised learning workflow from data collection through deployment

  4. Construct and interpret a simple decision tree, explaining information gain and entropy

  5. Summarize the k-Nearest Neighbors algorithm and when to prefer it over decision trees

  6. Evaluate a classifier using accuracy, precision, recall, and F1 score

  7. Explain overfitting and underfitting in terms of the bias-variance tradeoff

  8. Compare symbolic AI approaches (Units 3—​7) with machine learning on specific problem types

  9. Identify current AI trends including large language models and generative AI

Reading Assignments

Weekly Schedule

Day Activity Resource

Mon—​Tue

Sections 8.1 and 8.2: What is ML? + Supervised Learning

Berkeley CS 188 ML chapter

Wed

Section 8.3: Evaluating Models

scikit-learn docs

Thu

Sections 8.4 and 8.5: Connecting It All + Current Landscape

MAIEI State of AI Ethics

Fri—​Sun

Lab: Machine Learning Classifier (wine quality dataset)

Week8_Lab_Starter.ipynb

Get a head start with this introduction to machine learning.

Introduction to Machine Learning

Begin: 8.1 What Is Machine Learning? →

This work is licensed under CC BY-SA 4.0.