Projects supervised by Iain Murray

This page lists project dissertations authored by students who worked with my supervision for part of their undergraduate or master’s degree programme.

Past projects:

• Font Style Transfer using Deep Learning, UG5 project 2018.
• Recurrent Neural Networks for Dasher, UG4 project 2018.
• Benchmarking Differentiable Neural Computers, UG4 project 2018.
• In 2018 I collaborated with MSc supervisors from Amazon Development Centre Scotland, who took on 20 students, rather than supervising projects directly myself.
• A Model for predicting the Outcomes of Professional Tennis Matches, MSc 2017.
• Differentiating linear algebra routines, MSc 2017.
• Fast inference for Cosmology, MSc 2017.
• Probabilistic modelling for inventory forecasting, MSc 2017.
• When and why do differentiable deep learning systems lose to random forests?, MSc 2017.
• Differentiating Gaussian process models, MSc 2017.
• A Synthetic Data Generating Approach and Synthetic Data Mechanisms for Regularization, MSc 2017.
• What is the capacity of printed paper?, UG4 2016.
• Machine learning of fonts, UG4 2016.
• (I took a six-month sabbatical in Spring/Summer 2016.)
• Distilling Model Knowledge, George Papamakarios, MSc 2015.
• Approximate inference for large-scale probabilistic ranking, MSc 2015.
• Implementing a training procedure for Mixture Density Network with full covariance components, MSc 2015.
• Machine Learning Approaches to Automatic Validation of Fibre Placement in Carbon Fibre Composite Manufacturing, MSc 2015. (Supervised by Korin Richmond; I was a second supervisor.)
• Machine learning and inferring cosmological parameters, 2015.
• Spotting presentation slides in video, UG4 project, 2015.
• Machine learning and inferring cosmological parameters, MSc 2014.
• Machine learning of multivariate probabilities, MSc 2014.
• Realtime Recommendation System for Online Games, MSc 2014.
• Comparison of Markov chain Monte Carlo methods for Gaussian Processes, MSc 2014.
• Machine learning of dependencies in high density neural recordings, Pol Moreno Comellas, MSc 2013.
• Measuring pulse rate with a webcam, Abel Gonzalez Garcia, MSc 2013.
• Better approximate inference for probabilistic ranking, MSc 2013.
• Automatic evaluation of photographic images, UG4 project, 2013.
• Unshredding, Razvan Ranca, UG4 project, 2013.
  International Winner, Undergraduate Awards 2013, Computer Science.
• DOG removal — Inpainting Digital Onscreen Graphics, MSc 2012.
• Fast optimization of non-convex Machine Learning objectives, Nikolaos Nikolaou, MSc 2012.
• Improving the appearance and compression of scanned documents, MSc 2012.
• Effect of graph structure on predicting game outcomes, MSc 2012.
• Cleaning up photos of indexed colour documents, UG4 project, 2012.
• Fast low-rank metric learning, Dan Oneaţă, MSc 2011.
• Rating systems with multiple factors, Marius Stănescu, MSc 2011.
• Empirical evaluation of Gaussian process approximation algorithms, Krzysztof Chalupka, MInf 2011. (Supervised by Chris Williams; I was a second supervisor.)

University of Edinburgh Students: Sorry, I'm not supervising any projects this year. I'm managing MSc projects supervised by Amazon Development Centre Scotland, and am not involved in UG projects. I won't be taking on self-proposed projects.

I don't supervise project students from other Universities.

Any past student who would like links to their name or project added or removed, get back in touch.