Modelling and Classification of Data

Module code: EG7013

This module will present you with the fundaments of a wide range of algorithms (Bayes classifiers, neural networks and fuzzy logic) available to perform key artificial intelligence (AI) tasks, including machine learning, pattern classification, modelling and control. We also cover the fundaments of information theory with common examples for data quantification and storage. At the end of this module, you should be able to appreciate the rich variety of the existing techniques and be familiar with the literature. This module provides key tools for you, if you are interested in further research in the areas of AI and big data.

  • Learning to select a particular method from standard pattern recognition techniques such as linear discriminant functions, fuzzy and neural networks
  • Demonstrating an understanding on random variables and concepts from information theory, being able to fit a distribution to data collected in the field and other skills
  • Calculating error probability for a statistical classifier
  • Calculating optical decision boundaries for data classification problems and recognise different forms of pattern recognition problems such as classification and regression
  • Implementing Bayes classification methods which might be required in dealing with large volumes of data via computer tools

Learning

  • 20 hours of lectures
  • 2 hours of practicals
  • 128 hours of guided independent study

Assessment

  • Written exam, 2 hours (70%)
  • Computer exam (30%)