Machine Learning

1. 01
   Part 1ML foundations8 topics1 project

   • Types of ML: supervised, unsupervised, reinforcement
   • Maths you need: vectors, matrices, derivatives (simple)
   • NumPy and pandas refresher
   • The ML workflow from data to model
   • Train, validation and test sets
   • Bias and variance
   • Overfitting and underfitting
   • scikit-learn basics

   Project Your first model: predict student marks.

2. 02
   Part 2Supervised learning8 topics1 project

   • Linear regression and gradient descent
   • Logistic regression
   • k-nearest neighbours
   • Decision trees
   • Random forests
   • Gradient boosting: XGBoost and LightGBM
   • Support vector machines
   • Regression and classification metrics

   Project A loan approval or spam classifier.

3. 03
   Part 3Unsupervised learning8 topics1 project

   • k-means clustering
   • Hierarchical clustering
   • DBSCAN
   • PCA for dimensionality reduction
   • Anomaly detection
   • Customer segmentation project
   • Recommendation systems basics
   • Evaluating clusters

   Project Group shop customers into segments.

4. 04
   Part 4Better models8 topics1 project

   • Feature engineering and scaling
   • Encoding categories
   • Handling imbalanced data
   • Cross-validation
   • Hyperparameter tuning: grid and random search
   • Pipelines and ColumnTransformer
   • Model explainability with SHAP
   • Avoiding data leakage

   Project Improve a model and show the before and after.

5. 05
   Part 5Deploy8 topics1 project

   • Saving models with joblib
   • A prediction API with FastAPI
   • A Streamlit app
   • Docker basics
   • Deploying to the cloud
   • Monitoring model drift
   • Experiment tracking with MLflow
   • Portfolio project and interview prep

   Project A live price-prediction app.

This is the planned syllabus. Small changes are possible before launch.