Machine Learning: Classification Course Syllabus

Overview (80-120 words) describing structure and time commitment.

Module 1: Introduction to Classification

Estimated time: 1 hour

• Overview of classification and its real-world applications
• Understanding classification use cases in industry and research
• Introduction to tools and data used in the course

Module 2: Linear Classifiers and Logistic Regression

Estimated time: 3 hours

• Implement logistic regression from scratch
• Explore decision boundaries and linear separability
• Apply gradient ascent for parameter optimization
• Solve multi-class problems using one-vs-all classification

Module 3: Decision Trees

Estimated time: 3 hours

• Understand how decision trees split data by feature values
• Learn tree construction and recursive partitioning
• Apply stopping rules to control tree depth
• Prevent overfitting in decision tree models
• Apply decision trees to structured and unstructured data

Module 4: Boosting for Classification

Estimated time: 2 hours

• Introduction to ensemble learning concepts
• Learn boosting techniques to improve weak learners
• Build strong classifiers from ensembles of weak models

Module 5: Scaling With Stochastic Gradient Ascent

Estimated time: 2 hours

• Use stochastic gradient ascent for large-scale learning
• Handle massive datasets efficiently with incremental updates
• Learn convergence behavior and optimization strategies

Module 6: Handling Missing Data and Model Evaluation

Estimated time: 2 hours

• Apply techniques to manage incomplete input data
• Evaluate models using accuracy, precision, and recall
• Interpret performance with ROC curves

Prerequisites

• Familiarity with Python programming
• Basic understanding of linear algebra and calculus
• Background in data manipulation and analysis

What You'll Be Able to Do After

• Build and train logistic regression models for binary and multi-class tasks
• Construct and interpret decision trees for classification
• Improve model performance using boosting techniques
• Scale learning algorithms to large datasets using stochastic methods
• Evaluate classification models using precision, recall, and ROC analysis