Introduction to Machine Learning SJP (aka USJ Srilanka)

Course designed to teach the fundamentals of machine learning, from regression and classification to deployment and an introduction to deep learning.

Week 1

Topics

1. Introduction
2. What is Machine Learning (Layman's Term)
3. Rule Based vs ML Systems
4. Learning Paradigms
5. Machine Learning Applications
6. Setting up development environment

Week 2

Topics

1. Supervised Learning (A Deep Dive)
2. Machine Learning Life Cycle
3. Data Splits
4. Model Generalization
5. Bias-Variance Tradeoff
6. Cross validation
7. Linear Regression

Week 3

Topics

1. Exploratory Data Analysis (EDA)
   • Univariate Analysis
   • Bivariate Analysis
2. Feature Importance - Correlation
   • Heatmaps
3. Handling Categorical Features
   • One-Hot Encoding
4. House Price Prediction Example

Week 4

Topics

1. Logistic Regression
2. Evaluation Matrics
   • Accuracy
   • Precision
   • Recall
   • ROC Curves & AUC
   • PR Curves & AU-PRC
3. Dummy Models
4. Scaling Data
   • Standardization
   • Normalization
5. Practical Examples

Week 5

1. Decision Trees
   • CART
   • Gini Impurity & Gini Gain (aka Information Gain from Gini Impurity)
   • MAE Impurity & Variance Reduction
2. Overfitting
3. HyperParameter Tuning
4. Grid Search
5. Ensembling
   • Bagging (aka Bootstrapping)
   • Boosting
   • Stacking

Week 6

1. Introduction to Unsupervised Learning
2. Applications of Unsupervised Learning
3. KMeans Clustering
4. Within-Cluster Sum of Squares and Elbow plot
5. Evaluating Unsupervised Models
6. Practical Examples