Machine Learning and Data Science Foundation

How to work with real-world data — from messy, unstructured inputs through cleaning, exploratory analysis, and visualisation to extract meaningful stories and insights. How to build and evaluate ML models — covering core algorithms like regression, decision trees, SVMs, and ensemble methods like XGBoost and LightGBM, with rigorous model evaluation beyond just accuracy. How to engineer features and handle data challenges — including encoding, scaling, feature selection, imbalanced datasets, and proper train/test validation strategies that mirror industry practice. How to apply unsupervised learning and dimensionality reduction — using K-Means, DBSCAN, PCA, and t-SNE to find hidden patterns and compress high-dimensional data without losing signal. How to take a model from notebook to production — building reproducible ML pipelines, serving models with FastAPI and Docker, monitoring for data drift, and presenting your work as a portfolio-ready, industry-standard project.