Full Stack Data Science Course

Lesson 1 • Python Programming Essentials

Covers Python syntax, data structures, and control flow as the primary coding language. Establishes the scripting foundation all subsequent data work depends on.

Lesson 2 • Linear Algebra for Data Science

Teaches vectors, matrices, and decompositions as the mathematical backbone of ML algorithms. Connects directly to dimensionality reduction and model optimisation later.

Lesson 3 • NumPy for Numerical Computing

Introduces array-based computation using NumPy for fast numerical operations. Directly enables matrix maths and vectorised data transformations used throughout the course.

Lesson 4 • Calculus and Optimisation Basics

Covers derivatives, gradients, and the chain rule as tools for understanding model training. Provides the intuition behind gradient descent used in every ML chapter.

Lesson 1 • Data Visualisation with Matplotlib and Seaborn

Builds chart-creation skills using Matplotlib and Seaborn for communicating data insights. Visualisation fluency supports both EDA and stakeholder reporting throughout the course.

Lesson 2 • Data Ingestion and Pandas Basics

Introduces DataFrames, Series, and file I/O for loading structured data from multiple sources. Forms the entry point for all data manipulation tasks in this chapter.

Lesson 3 • Feature Engineering and Transformation

Teaches encoding, scaling, binning, and derived feature creation to prepare data for ML. Transformed features are the direct input to all supervised learning models.

Lesson 4 • Exploratory Data Analysis Techniques

Applies statistical summaries and visualisations to uncover patterns, correlations, and anomalies. EDA findings guide feature selection and model choice in subsequent chapters.

Lesson 1 • Core SQL Query Writing

Covers SELECT, WHERE, ORDER BY, and LIMIT clauses for retrieving and filtering data. These are the building blocks for every analytical query in this chapter.

Lesson 2 • Relational Database Concepts

Explains tables, keys, schemas, and normalisation as the structural foundation of relational databases. Understanding schema design is prerequisite to writing effective queries.

Lesson 3 • SQL for Data Science Workflows

Applies SQL to data extraction, feature computation, and pipeline integration with Python. Bridges database querying and the pandas-based analysis covered in Chapter 2.

Lesson 4 • Subqueries and Window Functions

Introduces subqueries, CTEs, and window functions for advanced analytical SQL patterns. These techniques are essential for ranking, running totals, and cohort analysis.

Lesson 1 • Hyperparameter Tuning and Cross-Validation

Covers grid search, random search, and k-fold cross-validation for robust model optimisation. These techniques prevent overfitting and ensure reliable performance estimates.

Lesson 2 • Machine Learning Workflow and Scikit-Learn

Introduces the end-to-end ML pipeline: splitting data, fitting models, and evaluating predictions. Establishes the scikit-learn API pattern used in every subsequent ML section.

Lesson 3 • Ensemble Methods and Boosting

Introduces random forests, gradient boosting, and XGBoost as high-performance ensemble models. Builds on single-model concepts to achieve state-of-the-art results on tabular data.

Lesson 4 • Regression Models and Evaluation

Covers linear, ridge, lasso, and polynomial regression for continuous target prediction. Evaluation metrics like RMSE and R² are introduced here and reused throughout the chapter.

Lesson 1 • Anomaly Detection Methods

Applies isolation forest, one-class SVM, and statistical methods to detect outliers in data. Anomaly detection is a direct application of unsupervised learning to real-world problems.

Lesson 2 • Nonlinear Dimensionality Reduction

Introduces t-SNE and UMAP for visualising high-dimensional data in two or three dimensions. These methods reveal cluster structure that PCA cannot capture in nonlinear data.

Lesson 3 • Topic Modelling and Matrix Factorisation

Covers NMF and LDA for extracting latent topics from document collections and sparse matrices. Extends unsupervised learning to text and recommendation system use cases.

Lesson 4 • Principal Component Analysis

Covers PCA as a linear technique for variance-preserving dimensionality reduction. Connects eigendecomposition from Chapter 1 to practical feature compression and visualisation.

Lesson 1 • Neural Network Fundamentals

Introduces perceptrons, activation functions, and backpropagation as the core mechanics of neural networks. Connects gradient descent from Chapter 1 to weight update rules.

Lesson 2 • Convolutional Neural Networks

Covers convolution, pooling, and CNN architectures for image classification and feature extraction. Builds on feedforward networks to handle spatial data structures.

Lesson 3 • Building Models with Keras and TensorFlow

Teaches the Sequential and Functional APIs for constructing and training deep learning models. Establishes the coding patterns used in all subsequent deep learning sections.

Lesson 4 • Recurrent Networks and Sequence Modelling

Introduces RNNs, LSTMs, and GRUs for modelling sequential and time-series data. Prepares students for NLP and forecasting tasks covered in later chapters.

Lesson 1 • Named Entity Recognition and Information Extraction

Applies sequence labelling models and spaCy to extract entities, relations, and structured data from text. Connects NLP models to practical information extraction pipelines.

Lesson 2 • Word Embeddings and Semantic Representations

Introduces Word2Vec, GloVe, and FastText for dense semantic word representations. Embeddings replace sparse vectors and improve model performance on downstream NLP tasks.

Lesson 3 • Transformer Models and BERT

Introduces the attention mechanism, transformer architecture, and BERT for contextual NLP. Covers fine-tuning pretrained models for classification and named entity recognition.

Lesson 4 • Text Preprocessing and Representation

Covers tokenisation, stemming, stop-word removal, and TF-IDF for converting raw text to features. These preprocessing steps are the foundation of every NLP pipeline in this chapter.

Lesson 1 • ML Pipelines and Workflow Orchestration

Introduces Airflow and Prefect for scheduling, orchestrating, and monitoring ML training pipelines. Automated pipelines replace manual retraining and reduce operational risk.

Lesson 2 • Model Monitoring and Drift Detection

Covers data drift, concept drift, and performance degradation monitoring for deployed models. Continuous monitoring closes the loop between deployment and model retraining.

Lesson 3 • ML Project Structure and Reproducibility

Establishes project layout, virtual environments, and version control practices for reproducible ML. Reproducibility is the prerequisite for every deployment and collaboration task in this chapter.

Lesson 4 • Experiment Tracking and Model Registry

Applies MLflow for logging parameters, metrics, and artefacts across training runs. A model registry enables versioned promotion from experiment to production.