Machine Learning

This course aims to introduce machine learning methods and their application to real-world problems. Supervised and unsupervised methods will be introduced, along with best practices for model evaluation and tuning. A data-centric approach to improving performance will also be discussed. The course focuses on implementing various machine learning models in Python and solving practical problems using ML libraries.

1. Introduction to Machine Learning (1 session)
2. Regression (4 sessions)
   • Linear regression
   • Gradient descent optimization
   • Nonlinear regression
   • Regularization
   • Overfitting
   • Validation and model selection
3. Classification (4 sessions)
   • Decision trees
   • Support Vector Machines (SVM)
   • Probabilistic classification: Naïve Bayes and Logistic regression
   • Instance-based learning methods such as kNN
4. Evaluation Metrics for Classification and Regression (1 session)
5. Ensemble Learning (2 sessions)
   • Bagging approach: Random Forest
   • Boosting approach: Boosted trees methods like xgBoost
6. Dimensionality Reduction (2 sessions)
   • PCA method
   • SVD method
7. Clustering (2 sessions)
   • K-means method
   • Hierarchical clustering
   • Clustering evaluation
8. Practical Problem-Solving Techniques (2 sessions)
   • Presentation of applied problems and solution pipelines

• Quiz: 20%
• Assignments: 40%
• Final Exam: 40%
• Pop Quizzes: 10%