MENU

You are here

Neural networks: how they work and their future in science

Tom Charnock
Location: 
A-005
Schedule: 
Wednesday, January 22, 2020 - 10:00
Abstract: 

I will present a short, methodological, introduction to neural networks - and paint as statistical models which can be used to approximate arbitrary functions. I will then show a few examples, concentrating on the current uses of neural networks from different areas of science before describing how these new data analysis tools can be used for optimally extracting information from data. In particular, using specific knowledge from the interdisciplinary communities of machine learning, statistics and informatics, I will show how we can use neural networks to inform us about physical processes and the correlations between the distribution of information in data and the models we use to describe the data.

Sign in