COMPGI01 - Supervised Learning

This database contains the 2017-18 versions of syllabuses.

Note: Whilst every effort is made to keep the syllabus and assessment records correct, the precise details must be checked with the lecturer(s).

Code COMPGI01 (Also taught as: COMPM055 Supervised Learning)
Year MSc
Prerequisites Basic mathematics, Calculus, Probability, Linear algebra
Term 1
Taught By Mark Herbster (50%), John Shawe-Taylor (30%), Massi Pontil (20%)
Aims This module covers supervised approaches to machine learning.
Learning Outcomes Gain in-depth familiarity with various classical and contemporary supervised learning algorithms, understand the underlying limitations and principles that govern learning algorithms and ways of assessing and improving their performance.


The course consists of both foundational topics for supervised learning such as Linear Regression, Nearest Neighbors and Kernelisation as well contemporary research areas such as multi-task learning and optimisation via proximal methods. In a given year topics will be drawn non-exclusively from the following.

  • Nearest Neighbors
  • Linear Regression
  • Kernels and Regularisation
  • Support Vector Machines
  • Gaussian Processes
  • Decision Trees
  • Ensemble Learning
  • Sparsity Methods
  • Multi-task Learning
  • Proximal Methods
  • Semi-supervised Learning
  • Neural Networks
  • Matrix Factorization
  • Online Learning
  • Statistical Learning Theory

Method of Instruction

Lecture presentations with associated class problems


The course has the following assessment components:

  • Written Examination (2.5 hours, 75%)
  • Coursework Section (25%)

To pass this course, students must:

  • Obtain an overall pass mark of 50% for all sections combined.

For full details of coursework see the course web page.


Reading list available via the UCL Library catalogue.