Machine Learning Research Topics And Application Fields

‘Machine learning gives you all the fun of getting involved in the analysis of amazing scientific data without the hassle of having to actually gather it. Machine learning is perhaps the emerging service industry of science and commerce.’ Professor and Director of the Computational Statistics and Machine Learning (CSML) John Shawe-Taylor, Department of Computer Science, University College London

An essential part of the MSc in Machine Learning is the five month master thesis. Your supervisor will assist, facilitate and then evaluate your master project. But the project requires you to organize your work independently, to submit it on time and take responsibility for it. Projects may range from very theoretical to very applied projects. You can choose either purely academical projects or you also may have an industrial collaborator. Especially for theoretical projects a good first pointer is to look at the recent publications of the members from the Centre for Computational Statistics and Machine Learning and the Gatsby Unit Research or at the challenges of the Pascal Network.

As stated UCL is the scientific coordinator of the Pattern Analysis, Statistical Modelling and Computational Learning (PASCAL) Network of Excellence. Within the Pascal network many machine learning research challenges are posed which are known as Pascal Challenges. These challenges and many other application fields can be combined with machine learning and form the basis for a master project. In the following account we give a sample of application fields students could be working on in their master thesis. Staff members are open to apply machine learning to new topics which are only limited by your imagination.

• General theoretical machine learning topics

• Collaborative filtering

• Behavioural targeting

• Machine Vision

• Music

• Language

• Information Retrieval

• Bioinformatics and computational biology

• Neuroscience

• Finance

• Digital Biology

• Inference in Complex Stochastic Environmental Models

• Applied Games