Machine Learning and Visual Analytics: Understanding Data by Eye
In today's digital world, how can we understand the vast quantities of data that surround us? This talk will show how machine learning can be used to uncover the hidden meanings of data and provide non-specialist users with the information they need in a form that they can use.
The key to doing this is through probabilistic models with latent (hidden)variables: this allows us to handle uncertainty in a rigorous way. In particular, the talk will describe the principles underlying data projection - tools for representing high-dimensional data in 2D and 3D -and illustrate how interaction with visualization plots improves understanding.
The value of this approach will be demonstrated by applications to problems in drug design and bioinformatics, biomedical engineering, helicopter condition monitoring, and oil exploration.