Phase 3: Machine Learning Foundations
Days 40–54 | 15 Lessons
Dive into supervised and unsupervised learning, neural networks, and NLP. Build and evaluate predictive models for business applications.
Lessons in This Phase
| Day | Title | Description |
|---|---|---|
| 40 | Day 40: Introduction to Machine Learning & Core Concepts | This lesson introduces the machine learning workflow and highlights how evaluation techniques ensure |
| 41 | Day 41 · Supervised Learning – Regression | - solutions.py – modular helpers for generating synthetic regression data, training a linear |
| 42 | Day 42 · Supervised Learning – Classification (Part 1) | - solutions.py – reusable helpers for loading the Iris dataset, scaling features, and training |
| 43 | Day 43 · Supervised Learning – Classification (Part 2) | - solutions.py – modular helpers for preparing the Iris dataset, fitting SVM and decision tree |
| 44 | Day 44: Unsupervised Learning | Day 44 introduces two foundational unsupervised learning workflows: |
| 45 | Day 45: Feature Engineering & Model Evaluation | Day 45 demonstrates how thoughtful preprocessing and rigorous evaluation combine to build |
| 46 | Day 46: Introduction to Neural Networks & Frameworks | Welcome to Day 46! Today, we begin our exploration of Deep Learning by introducing the |
| 47 | Day 47: Convolutional Neural Networks (CNNs) for Computer Vision | Welcome to Day 47! Today, we dive into Convolutional Neural Networks (CNNs), a specialized type |
| 48 | Day 48: Recurrent Neural Networks (RNNs) for Sequence Data | Welcome to Day 48! Today, we explore Recurrent Neural Networks (RNNs), a class of neural |
| 49 | Day 49: Natural Language Processing (NLP) | Day 49 introduces the classic feature-extraction techniques that turn raw text into numeric |
| 50 | Day 50: MLOps - Model Deployment | Welcome to our final day, Day 50! Today, we touch upon MLOps (Machine Learning Operations), |
| 51 | Day 51 – Regularised Models | This lesson extends the regression toolkit with L2 (ridge), L1 (lasso), and elastic net penalties |
| 52 | Day 52 – Ensemble Methods | Day 52 highlights how bagging, boosting, and stacking unlock better accuracy than single estimators. |
| 53 | Day 53 – Model Tuning and Feature Selection | Optimisation is the bridge between baseline models and production-grade performance. Day 53 |
| 54 | Day 54 – Probabilistic Modeling | Gaussian mixtures, Bayesian classifiers, expectation-maximisation, and hidden Markov models power |