Data Science Engineering Course

Lesson 1 • Data Types and Storage Formats

Covers structured, semi-structured, and unstructured data formats and their storage systems. Prepares students to ingest diverse data sources.

Lesson 2 • Mathematics and Statistics Refresher

Reviews linear algebra, calculus, and probability essential for modelling. Provides the quantitative foundation required for all subsequent chapters.

Lesson 3 • Python for Data Science

Introduces Python syntax, data structures, and scientific libraries. Enables students to write reproducible analysis scripts from day one.

Lesson 1 • Data Transformation and Reshaping

Applies aggregation, pivoting, merging, and encoding to restructure data for analysis. Bridges raw data to feature-ready formats.

Lesson 2 • Exploratory Data Analysis

Teaches systematic profiling of datasets to detect distributions, outliers, and relationships. Drives informed decisions about cleaning and feature engineering.

Lesson 3 • Data Collection Methods

Covers web scraping, API calls, and database queries as primary data acquisition strategies. Connects data sourcing to downstream pipeline design.

Lesson 4 • Data Cleaning and Imputation

Addresses missing values, duplicates, inconsistent formats, and erroneous entries. Ensures data integrity before modelling begins.

Lesson 1 • A/B Testing and Experimentation

Designs controlled experiments, calculates sample sizes, and interprets results correctly. Directly applicable to product and business analytics roles.

Lesson 2 • Bayesian Inference Fundamentals

Introduces prior and posterior distributions and Bayesian updating as an alternative inference paradigm. Prepares students for probabilistic modelling in later chapters.

Lesson 3 • Hypothesis Testing Framework

Covers null and alternative hypotheses, p-values, and error types for rigorous testing. Connects statistical significance to practical decision-making.

Lesson 1 • Classification Models

Trains logistic regression, decision trees, and support vector machines for categorical targets. Addresses class imbalance and threshold selection.

Lesson 2 • Machine Learning Fundamentals

Defines supervised learning, the bias-variance tradeoff, and the train-test split paradigm. Establishes the conceptual framework for all modelling chapters.

Lesson 3 • Regression Models

Covers linear, polynomial, and regularised regression for continuous target prediction. Builds intuition for model coefficients and regularisation penalties.

Lesson 4 • Hyperparameter Tuning and Model Selection

Applies grid search, random search, and Bayesian optimisation to maximise model performance. Teaches principled model comparison and selection.

Lesson 1 • Association Rule Learning

Mines frequent itemsets and association rules to reveal co-occurrence patterns. Supports recommendation and market basket analysis applications.

Lesson 2 • Dimensionality Reduction Techniques

Applies PCA, t-SNE, and UMAP to reduce feature space whilst preserving structure. Enables visualisation and speeds up downstream modelling.

Lesson 3 • Anomaly Detection

Identifies outliers and rare events using statistical and model-based approaches. Applies directly to fraud detection and quality control scenarios.

Lesson 1 • Feature selection methods

Applies filter, wrapper, and embedded methods to identify the most predictive features. Reduces dimensionality and prevents overfitting.

Lesson 2 • Fairness and bias auditing

Detects and mitigates demographic bias in model predictions using fairness metrics. Ensures models meet ethical and organisational standards.

Lesson 3 • Model-agnostic interpretability

Uses SHAP and LIME to explain individual predictions and global model behaviour. Builds stakeholder trust and supports regulatory compliance.

Lesson 1 • Neural network fundamentals

Explains perceptrons, activation functions, forward propagation, and backpropagation. Provides the mathematical foundation for all deep learning architectures.

Lesson 2 • Recurrent neural networks and sequences

Trains RNNs, LSTMs, and GRUs for sequential and time-series data. Connects sequence modelling to NLP and forecasting applications.

Lesson 3 • Transformer architecture and attention

Introduces self-attention, multi-head attention, and the transformer block as the backbone of modern NLP. Prepares students for fine-tuning large language models.

Lesson 4 • Training optimisation and regularisation

Applies batch normalisation, dropout, learning rate scheduling, and early stopping to stabilise training. Reduces overfitting in deep models.

Lesson 1 • Model monitoring and drift detection

Tracks prediction quality, data drift, and concept drift in production to maintain model reliability. Triggers alerts and retraining when performance degrades.

Lesson 2 • CI/CD for machine learning

Automates testing, validation, and software deployment of ML pipelines using continuous integration practices. Reduces manual errors and accelerates release cycles.

Lesson 3 • Scalable ML infrastructure

Designs distributed training and inference infrastructure using cloud-native and orchestration tools. Prepares students for enterprise-scale ML system design.

Lesson 4 • Experiment tracking and model registry

Logs parameters, metrics, and artefacts using experiment tracking tools and model registries. Enables reproducibility and team collaboration on model development.