°ÄÃÅÁùºÏ²Ê¿ª½±¼Ç¼

Internal

CSMML16 - Machine Learning

°ÄÃÅÁùºÏ²Ê¿ª½±¼Ç¼

CSMML16-Machine Learning

Module Provider: Computer Science
Number of credits: 10 [5 ECTS credits]
Level:7
Terms in which taught: Spring term module
Pre-requisites:
Non-modular pre-requisites:
Co-requisites:
Modules excluded:
Current from: 2019/0

Module Convenor: Dr Tom Thorne

Email: t.thorne@reading.ac.uk

Type of module:

Summary module description:

This module covers the topic of machine learning.


Aims:

The aim of the module is to introduce students to current methods in machine learning and their application to real world problems.



Ìý


Assessable learning outcomes:

Students will be able to:




  • Explain support vector machines and k-means methods

  • Determine appropriate machine learning methods for clustering, classification and regression problems.

  • Apply machine learning methods to perform clustering, classification and regression.

  • Explain the process of training and making predictions with a neural network.

  • Determine the appropriate neural network architecture for a particular problem.

  • Apply multiple classes of neural network to real world problems involving image and text data.

  • Understand ensemble methods in machine learning.

  • Apply ensemble methods to real world data.


Additional outcomes:

Students will gain familiarity with machine learning and neural network libraries, and the Python programming language.


Outline content:

The module covers foundational topics in relevant machine learning algorithms:



Support Vector Machines



K-means clustering



Neural networks:




  • Backpropagation

  • Stochastic gradient descent

  • Activation functions

  • Feedforward and recurrent architectures

  • Convolutional neural networks

  • Generative adversarial networks

  • Capsule networks



Ensemble methods:




  • Boosting

  • Bagging

  • Stacking



Students will learn how to apply these methods in various domains using the Python language and libraries, including:




  • Image classification

  • Image synthesis

  • Natural language processing


Brief description of teaching and learning methods:

The module consists of 10 lectures and weekly guided practical classes that implement methods covered in the lectures.


Contact hours:
Ìý Autumn Spring Summer
Lectures 10
Practicals classes and workshops 10
Guided independent study: 80
Ìý Ìý Ìý Ìý
Total hours by term 100
Ìý Ìý Ìý Ìý
Total hours for module 100

Summative Assessment Methods:
Method Percentage
Written exam 50
Project output other than dissertation 50

Summative assessment- Examinations:

One 1.5 hour exam.


Summative assessment- Coursework and in-class tests:

One project based assignment, due week 11 of Spring term.


Formative assessment methods:

Feedback in practical classes.


Penalties for late submission:
Penalties for late submission on this module are in accordance with the University policy. Please refer to page 5 of the Postgraduate Guide to Assessment for further information: http://www.reading.ac.uk/internal/exams/student/exa-guidePG.aspx

Assessment requirements for a pass:

A mark of 50% overall.


Reassessment arrangements:

ÌýÌýOne examination paper of 2 hours duration in August/September.


Additional Costs (specified where applicable):

Last updated: 10 April 2019

THE INFORMATION CONTAINED IN THIS MODULE DESCRIPTION DOES NOT FORM ANY PART OF A STUDENT'S CONTRACT.

Things to do now