This module addresses the following topics:

Introduction and positioning of machine learning
- Role of ML within AI and data-driven systems (with minimal overlap with Level 4 introductions).
- Supervised vs unsupervised learning; basic terminology and workflow.

Data preparation and feature engineering
- Handling missing data, outliers, scaling and normalisation.
- Feature encoding for categorical variables; feature selection basics.
- Train/validation/test splits; cross-validation strategies.

Core supervised learning algorithms
- Linear and logistic regression (with regularisation: L1/L2).
- k-Nearest Neighbours.
- Decision trees and ensemble methods (Random Forests, Gradient Boosting at an introductory level).
- Support Vector Machines (conceptual and applied level appropriate for Level 5).

Core unsupervised learning algorithms
- Clustering: k-means, hierarchical clustering at an applied level.
- Dimensionality reduction: PCA and related techniques (conceptual, with practical examples).

Model evaluation and selection
- Evaluation metrics for regression (Ex. MSE, MAE, R²) and classification (accuracy, precision, recall, F1, ROC-AUC, and confusion matrix).
- Cross-validation and resampling; hyper-parameter tuning (Ex. grid/random search).
- Bias-variance trade-off and overfitting/underfitting.
- Handling imbalanced datasets (Ex. resampling strategies, appropriate metrics).

Interpretation, robustness and basic responsible ML concerns
- Basic model interpretability (Ex. feature importance, simple model explainability) without overlapping deeply with Human-Centred AI modules.
- Sensitivity to data quality and distribution shift.
- High-level awareness of fairness and bias issues (technical focus; detailed ethics/regulation covered elsewhere).

Experiment management and reproducibility
- Structuring ML projects; use of notebooks/scripts and version control.
- Recording experiments, hyper-parameters and results for reproducibility.

Applied case studies and employability
- Applied ML case studies from business, science, and technology domains (chosen so as not to overlap with Level 6 Applied AI: Vision, Language & Secure Systems).
- Emphasis on problem-solving, evaluation of systems, and communication in a professional context.

BCS / TechSkills elements:

Maths principles in ML algorithms and evaluation.
Problem solving and system modelling (choosing and configuring appropriate algorithms for given problems).
Evaluation of systems through metrics and diagnostics.
Legal, social, ethical and professional issues briefly (particularly around data quality and bias), with more in-depth coverage located in Human-Centred AI modules.

1. PRACTICAL - Practical Machine Learning Portfolio (40%)

You will complete a series of guided but progressively open-ended exercises in which you:

Implement several core machine learning algorithms (Ex. linear/logistic regression, k-NN, decision trees, ensemble methods, clustering) using Python and appropriate libraries (Ex. scikit-learn).
Construct and compare alternative pipelines for a given applied problem (Ex. classification or regression on tabular data).
Document their approach, configuration choices, and key performance results in short commentaries embedded within notebooks or a concise written summary.

2. REPORT - Individual Technical Report: ML Experiment Design and Evaluation (60%)

You will evaluate an end-to-end ML solution to a selected problem using a real-world dataset. The report will include:

Problem definition and justification of chosen target metric(s).
Rationale for data preparation, feature engineering and algorithm selection.
Rigorous evaluation using appropriate metrics, cross-validation and diagnostics, including discussion of overfitting, bias/variance trade-off and data limitations.
Interpretation of results and recommendations for deployment or further work, with reflection on professional implications (Ex. reliability, fairness at a high level, suitability for stakeholders).
Reflect on Machine Learning design decisions you have taken.

All teaching sessions will blend theory and practical learning:

Lectures to introduce core concepts, algorithms and evaluation methods.
Practical classes and workshops in computer labs to implement ML pipelines using Python and industry-standard libraries (Ex. scikit-learn).
Tutorials / drop-in sessions for guided support on lab work and assessment tasks.
Group discussions and short problem-solving activities to interpret results and share approaches.
Formative exercises (Ex. short coding challenges, low-stakes quizzes, mini-experiments) to build confidence and prepare for summative assessments.

You will also be provided with curated independent study resources (Ex. documentation, example notebooks, academic/industry articles) to deepen understanding and support self-directed learning.

1. Evaluate machine learning algorithms and their applications.

Knowledge & Understanding
Application & Problem-Solving

2. Design and implement end-to-end machine learning pipelines using appropriate tools.

Digital Literacy
Personal Development & Entrepreneurship

3. Evaluate the performance, robustness and limitations of machine learning models using appropriate metrics.

Critical Reasoning & Collaboration
Research Skills

4. Critically reflect on machine learning design decisions and results for technical and non-technical stakeholders.

Communication
Reflection

Students will require access to:

Networked computer labs with Python and Jupyter (or equivalent) installed.

Machine learning libraries such as scikit-learn, pandas, NumPy, and matplotlib.

Version control tools (Ex. Git, GitHub/GitLab) for managing code and experiments.

University VLE (Blackboard) for module materials, announcements and submissions.

Appropriate datasets (open-source or provided case-study datasets) hosted via the VLE or approved platforms.

Géron, A. (2022) Hands-On Machine Learning with Scikit-Learn, Keras & TensorFlow, 3rd edn. O’Reilly.

Brasil, J. (2025) Now Machine Learning Vol1: Supervised Learning: Mastering Regression, Classification, and Predictive Modelling with Python, Now Machine Learning

Scikit-learn (ongoing) “User Guide & API Reference” [Online] Available at: https://scikit-learn.org/

Parikha, D. (2025) Machine Learning Essentials You Always Wanted to Know: A Hands-On Beginner's Guide to Mastering AI, Supervised, Unsupervised, and Deep Learning Algorithms (Self-Learning Management Series), Vibrant Publishers

In this module, you will learn how to build and evaluate practical machine learning solutions to real-world problems. You will explore a range of core supervised and unsupervised algorithms, learn how to prepare and model data using contemporary tools, and develop the skills needed to compare, tune and justify different approaches. Through a combination of hands-on labs and an individual project, you will design complete machine learning pipelines, critically evaluate model performance, and communicate your findings in a way that aligns with professional AI and data science practice.