Andrew McDonald

I graduated from UCL with a PhD in machine learning in 2017. As of January 2018 I am a quantitative researcher at a London hedge fund. Prior to commencing my PhD I was a trader in structured products linked to alternative investments for eight years at KBC Financial Products, formerly part of D.E. Shaw.

Research Interests

My supervisor is Massimiliano Pontil, and I am funded by an EPSRC studentship. My research interests include:

• Structured sparsity and regularization

• Convex optimization

• Matrix completion and multitask learning

Publications

• Learning with Optimal Interpolation Norms (with P. Combettes, C.A. Micchelli, M. Pontil, journal paper under review)

• New Perspectives on k-Support and Cluster Norms (with M. Pontil, D. Stamos, JMLR 2016)

• Fitting Sparsity and Spectral Decay with the k-Support Norm (with M. Pontil, D. Stamos, AISTATS 2016)

• Learning with Dataset Bias in Latent Subcategory Models (with D. Stamos, S. Martelli, M. Nabi, V. Murino, M. Pontil, CVPR 2015)

• Regularization with the Spectral k-Support Norm (with M. Pontil, D. Stamos, NIPS 2014)

Contact

a.mcdonald at cs.ucl.ac.uk
My LinkedIn profile