Global Opportunities

CSC3831 : Predictive Analytics, Computer Vision & AI

Semester 1 Credit Value: 20
ECTS Credits: 10.0


This module aims to provide a foundation in the field data analytics, predictive data science, computer vision methods and machine learning.

Outline Of Syllabus

Selected topics chosen from:

Principles and practice of data analytics:
Fundamental data representations.
Data structures and data management infrastructure to enable data analytics.
Methods for data preparation, including source selection and integration, data cleaning.

Principles and practice of computer vision:
Background, Image Model, Spatial Coordinate, Digitisation.
Image Sampling, Image Quality, Image Pixel Relationships.
Linear Operators, 2-D Transforms. Spatial Domain Methods, Frequency Domain Methods.
Image Compression, Redundancy Types, Lossless and Lossy Compression, Compression Standards.
Object Detection Methods, Edge Liking and Boundary Detection.
Thresholding Methods, Region Oriented Methods. Pattern Recognition.
Mathematical Morphology

Principles and practice of machine learning and deep neural networks:
Covering core concepts, and techniques along with current leading approaches in machine/deep learning.

Teaching Methods

Teaching Activities
Category Activity Number Length Student Hours Comment
Scheduled Learning And Teaching ActivitiesLecture221:0022:00Lecture material pre-recorded. Lectures in person (PIP) and where possible also streamed live online
Scheduled Learning And Teaching ActivitiesPractical113:0033:00Practicals 1x3 hour drop in practical per week. PIP mode.
Guided Independent StudyProject work661:0066:00Practical coursework and portfolio preparation
Guided Independent StudyProject work51:005:00Reflective report preparation
Guided Independent StudyIndependent study221:0022:00Lecture follow-up
Guided Independent StudyIndependent study521:0052:00Background reading, guided reading, one article, chapter or equivalent per two weeks.
Teaching Rationale And Relationship

The teaching methods combine traditional lectures with practical sessions so that students can explore the topics covered in both a theoretical and practical context. Lectures outline the underlying principles, algorithms and theory, while practical lab work encourages students to implement the algorithms using rea-world data, in terms of applying the methods to real world data examples.
Lecture material will be pre-recorded and students have the opportunity to watch the videos ahead of the lecture. Lectures in person and where possible streamed live with recap will be available. Lecture follow up will consist of Q&A about the lecture material.

Assessment Methods

The format of resits will be determined by the Board of Examiners

Other Assessment
Description Semester When Set Percentage Comment
Portfolio1M100A series of programming exercises covering each element of the syllabus and contributing to a portfolio of evidence of understanding
Formative Assessments
Description Semester When Set Comment
Practical/lab report1MFormative assessment: a live demo session where students discuss and critically assess their portfolio with the module leaders
Assessment Rationale And Relationship

The assessment is based on case studies, using real world data, allowing students to explore practical application of the techniques and algorithms that have been learned. The reflective report offers students the opportunity to draw together the overall learning experience on machine learning and predictive analysis of data sets.

Reading Lists