Data Science School

Sunday 29th March 2020

2:00 - 3:00pm: Registration and Coffee

3:00 - 3:15pm: Welcome and Introduction

3:15 - 3:40pm: Lord Clement-Jones, 'Is Democracy Falling Behind Technology?'

Abstract: The last few years has seen the advent of ubiquitous and addictive social media and “surveillance capitalism” with personal data capture allied to algorithm driven targeted political advertising. As a result fake news is rife. All this potentially allows electorates to be unwittingly manipulated. Is our democracy at risk as a result? What is the solution? Do policy makers have the understanding, ability or appetite to respond and regulate appropriately?

3:45 - 5:15pm: Shmuel Weinberger, 'Mathematical models of social choice'

Abstract: Deciding what society wants, i.e. aggregating voters' choices to form a societal choice is a fundamental problem -- independent of the problem of finding mechanisms to achieve this. In this talk, I will discuss some insights that can be gleaned from a geometric perspective.

5:30 - 7:00pm: Mehrnoosh Sadrzadeh, 'Formal and Distributional Natural Language Processing'

7:15pm: Dinner

Monday 30th March 2020

Monday 30 March

8:15am: Breakfast

9:00 - 10:45am: Norman Fenton, 'Bayesian networks: combining data and knowledge'

Abstract: While machine learning algorithms in conjunction with ‘big data’ can provide automated ‘human like’ behaviour for some decision problems (such image recognition), true ‘artificial intelligence’ cannot be achieved without expert causal knowledge. This talk will explain how Bayesian networks – as causal probabilistic models of risk developed by a ‘smart data’ approach – can provide powerful decision-support and accurate predictions for problems where purely data driven methods fail. The ‘smart data’ approach combines data (often minimal) with expert knowledge. The talk will provide examples in chronic diseases, forensics, terrorist threat analysis, and even sports betting.

11.00 - 1:00pm: William Marsh, 'Health AI'

Abstract: Does a successful decision-support system just require mastery of prediction technologies? We will consider the challenges of choosing an appropriate problem, eliciting medical knowledge, designing a usable system and generating evidence of its performance.

1pm: Lunch

2.00 - 4.00pm: Gilles Zemor, 'Introduction to Quantum Computing'

Abstract: The purpose of this talk is to give a self-contained introduction to the mathematics of quantum computing. We will explain how quantum superposition is used to instantiate some degree of parallel computing which yields an asymptotic speedup on classical computers. The focus will be on the algorithmic techniques that lead up Shor's integer factoring algorithm.

4.30 - 6.30pm: Gunnar Carlsson, 'Introduction to topological data analysis'

7:15pm: Dinner

Tuesday 31st March 2020

Tuesday 31 March

8:15am: Breakfast

9:00 - 10:45am: Gilles Zemor, 'Quantum Coding'

Abstract: Quantum processors are extremely fragile and the big challenge of quantum computing is to preserve the integrity of quantum states. Solutions are almost sure to involve quantum error-correcting codes in some way. We will give an introduction to the theory of quantum error-correcting codes, and highlight the LDPC (Low Density Parity Check) variety, which is arguably the most likely to be eventually used by quantum computers. We shall survey some achievements and challenges.

11.00 - 1:00pm: Jens Eisert, 'Tensor networks, quantum game theory'

1pm: Lunch

2.00 - 4.00pm: Mike Barnes with David Watson, 'Machine learning and AI in Precision Medicine'

4.30 - 6.30pm: Trevor Graham with Benjamin Werner, 'Mathematical modelling of cancer'

7:15pm: Dinner