Gaussian processes.Theory and applications in predictive modeling of spatiotemporal phenomena.

Example samples of a GP

In this report we present a tutorial on Gaussian Processes. The report explainsthe theory behind them, provide implementation details and conduct series ofexperiments to give a further intuition of how they work. We touch on number ofmore advance topics like connections with neural nets and hierarchical GaussianProcesses. In the second part of the report we look into the problems of activesensing and modeling spatiotemporal phenomena and why Gaussian Processesare well suited for those tasks.

Martin Asenov
PhD Candidate

My research interests include robotics, machine learning, computer graphics and the interplay between them.