To give students a grounding in foundational elements of data science theory and systems, including Data transformations, Database Systems, and practical data processing pipelines.

TBC

None

Programming and Systems Development (H)

Exam worth 60%

Weekly assessed in-lab exercises worth 40%

This module provides a solid foundation in data science theory and practice, focusing on:

Core Foundations:

■ Linear Algebra: Essential for data representation and transformation, useful in machine learning algorithms as well as in other domains.

■ Probability Theory: Understanding uncertainty, probabilistic models, and statistical reasoning.

■ Optimisation: Training methods for large-scale machine learning and deep learning/AI models, including gradient-based approaches and performance considerations.

■ Data Engineering & Lifecycle: Managing data from acquisition through storage, integration, and analysis.

Applied Components:

■ Data Transformation: Array manipulation, linear algebra routines, visualisation, and probabilistic concepts

■ Data Stores & Ecosystems: Principles of data store design and selection, and their role in data-intensive systems within real-world data science scenarios

■ Data Processing Pipelines: Practical workflows indicatively using Pandas, Numpy, Scikit-learn for cleaning, integrating, and analysing diverse data; including text representation, vectorisation, similarity measures, clustering, linking, and visualisation

By the end of this course students will be able to: 

1. formulate problems in tensor form, manipulate tensors and run vectorised code efficiently;

2. vectorise data, measure distances between items and apply matrix transformation and decomposition to practical problems;

3. formulate and understand problems with stochastic elements;

4. formulate problems as optimisation and select appropriate approaches to solve them;

5. understand and appreciate data lifecycle principles and data-intensive best practices for designing scalable, efficient solutions to large-scale data science challenges;

6. Select appropriate data storage strategies based on specific pre-processing requirements and the velocity of incoming data;

7. Apply governance and data provenance principles.