Statistics of Big Data and Machine Learning

School of Mathematics and School of Medicine are organising a three-day event devoted to statistical methods in big data, machine learning and applications in genomics.
The sponsor of this initiative is Data Innovation Research Institute of Cardiff University.

Organizing committee: Anatoly Zhigljavsky, Valentina Escott-Price, Alexander Balinsky, Jon Gillard and Andrey Pepelyshev.

Abstracts of talks are available here.

Programme:

November 6, 2018

• Venue: Council Chamber, Main Building (Main/1.77)
  
  

  • 13.45-14.00   Opening
  • 14.00-14.50 Lars Carlsson Conformal Prediction in Machine Learning and Applications
  • 14.50-15.40 Alexander Gammerman Reliable Machine Learning by Conformal Predictors

  
  
• Venue: Wallace Lecture Theatre, Main Building (Main/0.13)
  
  

  • 16.00-16.50 Coralia Cartis Stochastic variants of classical optimization methods, with complexity guarantees
  • 16.50-18.20 Arthur Gretton The Maximum Mean Discrepancy for Training Generative Adversarial Networks

  
  

• 09:00-13:00   Tutorials on Machine Learning. Venue: Committee Room 1, Glamorgan Building
  
  

  • 09.00-09.50 Alexander Balinsky Mathematics of Deep Learning
  • 10.00-10.50 Irena Spasić Multi-word Expressions in Text Mining
  • 10.50-11.10   coffee break
  • 11.10-12.00 Francesco Archetti Optimization and Machine Learning and Antonio Candelieri Optimization and Machine Learning with Applications
  • 12.10-13.00 Timos Kipouros Visual Analytics for High-Dimensional Data Exploration and Engineering Design Optimisation

  
  
• 13:00-14:00   lunch
  
  
• 14:00-14:50 Anatoly Zhigljavsky Peculiarities of High Dimensions. Venue: 1.64 lecture theatre, Glamorgan Building
  
  
• 14:50-17:30   Workshop on Big Data Issues in Genomics. Venue: 1.64 lecture theatre, Glamorgan Building
  
  

  • 14.50-15.40 Valentina Escott-Price Introduction to genomic data, overview of machine learning attempts to analyse genomic data
  • 15.45-16.05 Dobril Ivanov Using Genomic Data to Describe Phenotypes
  • 16.10-16.30 Matthew Smith Genomic data analyses with different machine learning approaches
  • 16.35-16.55 Elliott Rees Rare variation in human genome, next generation sequencing data generation, processing and current state of analyses
  • 17.00-17.20 Emily Baker POLARIS: Polygenic LD-Adjusted Risk Score Approach for Set-Based Analysis of Genomic Data

November 8, 2018:

• 09.30-10.00 Coffee and networking
• 10.00-12.30 Roundtable Discussion on Big Data and Industrial Applications. Venue: Committee Room 1, Glamorgan Building
  Several 10-15 min presentations followed by discussions.
  Confirmed speakers:

  • Prof Francesco Archetti Slides (University of Milano-Bicocca & Consorzio Milano Ricerche)
  • Prof Antonio Candelieri Slides (University of Milano-Bicocca & Optimization Analytics Knowledge Simulation)
  • Dr Alistair Edwardes (Office for National Statistics)
  • Dr Angelico Fetta (McLaren)
  • Dr Roman Guchenko (CrimTan)
  • Dr Penny Holborn (University of South Wales / Tata Steel)
  • Dr Matthew Jones (Nationwide)
  • Dr Leanne Smith (London Ambulance Service)