The launch for the new RSS section for Machine Learning and Computational Statistics was held on the 25th of January 2018. The launch event consisted of a series of talks giving introductions to machine learning, work into methodological issues, and examples of applications to real world issues.
The launch started with a talk from Sylvia Richardson of the MRC Biostatistics Unit, Cambridge. She gave a talk covering work on data compression with statistical guarantees, specifically methods to conduct multivariate regression and model exploration in datasets that can contain upwards of 500000 individuals.
The second talk was given by Zoubin Ghahramani of the University of Cambridge and Uber AI labs. He gave a clear introduction to the areas of machine learning and computational statistics, including explanations of various key terms. His talk focussed on the area of deep learning.
The launch continued with a talk from Julien Cornebise of Element AI, who described ongoing work using machine learning to assemble evidence of destruction of villages in Sudan through examination of satellite photographs of the area.
The final talk was delivered by Yee Whye Teh of the University of Oxford and Deepmind. He described the benefit of using Bayesian approaches, such as utilising prior information, for complex networks of data (such as where there is a large amount of data, but relatively little data for each individual in the analysis).
The launch concluded with a networking session, which allowed researchers with a background or interest in the area to connect.