BAYESIAN STATISTICAL MODELLING AND ANALYSIS IN SPORT

ABS Schools

The Applied Bayesian Statistics summer school has been running since 2004. Since 2012 it is organized by

• IMATI CNR Istituto di Matematica Applicata e Tecnologie Informatiche, Consiglio Nazionale delle Ricerche
• Dipartimento di Scienze Statistiche Università Cattolica, Milano

Since 2014 the school is organized in cooperation with Fondazione "Alessandro Volta"

The school aims to present state-of-the-art Bayesian applications, inviting leading experts in their field.
Each year a different topic is chosen.

Past editions were devoted to: Gene Expression Genomics, Decision Modelling in Health Care, Spatial Data in Environmental and Health Sciences, Bayesian Methods and Econometrics, Bayesian Decision Problems in Biostatistics and Clinical Trials, Bayesian Methodology for Clustering, Classification and Categorical Data Analysis, Bayesian Machine Learning with Biomedical Applications, Hierarchical Modeling for Environmental Processes, Stochastic Modelling for Systems Biology, Bayesian Methods for Variable Selection with Applications to High-Dimensional Data and Applied Bayesian Nonparametrics, Modern Bayesian Methods and Computing for the Social Sciences, Bayes Big Data and the Internet, Modeling Spatial And Spatio-Temporal Data With Environmental Applications.

The lecturer is: Prof. Kerrie Mengersen, Queensland University of Technology, Brisbane, Australia.

The topic chosen for the 2018 school is
Bayesian Statistical Modelling and Analysis in Sport.

Distinguished Professor Kerrie Mengersen is the past ISBA (International Society for Bayesian Analysis) President. She is the author of over 200 refereed journal publications, supervisor of over 30 postgraduate students in the past 5 years, recipient of over 30 large research grants. In 2016 she was awarded the Pitman Medal, the highest honour to be presented by the Statistical Society of Australia and the first woman to receive it. More details on her outstanding career can be found at her webpage.

She will be assisted by Dr. Paul Wu (Queensland University of Technology, Brisbane, Australia) for the practical sessions.

The aim of this course is to increase students' ability to develop Bayesian models and computational solutions for real problems in the world of sport. A case study based teaching approach will be taken for the course. Each day, students will be presented with one or two problems posed by Sports Institutes regarding aspects of athlete training for world games. Through participatory problem solving, the students will be challenged to learn about theory, methods and applications of a range of Bayesian models including mixtures, spatio-temporal models, hidden Markov models and experimental design, and computational approaches including Markov chain Monte Carlo and Approximate Bayesian Computation. This hands-on course pays equivalent attention to theory and application, foundation and frontiers in Bayesian modelling and analysis. While the focus of the case studies is on sport, both sporting novices and lovers of sports are welcome, noting that the learning obtained in the course will be widely applicable to many other areas.
Detailed programme

• Day 1: Lectures on introduction to Bayesian modelling and computation. Presentation of Problem 1: ranking and benchmarking athletes. Discussion and implementation of potential Bayesian hierarchical models and computational solutions. Communication of results.
• Day 2: Lectures on foundational Bayesian theory. Presentation of Problem 3: modelling swimmers' effective work per stroke. Discussion and implementation of potential Bayesian high dimensional regression models and computational solutions. Communication of results. Presentation of Problem 4: modelling cyclists' wearable data. Discussion and implementation of potential (marked) time series models and computational solutions. Communication of results.
• Day 3: Lectures on foundational Bayesian computation. Presentation of Problem 5: optimising athletes' resilience. Discussion and implementation of potential Bayesian mixture models to relate performance, fatigue and recovery. Communication of results.
• Day 4: Lectures on foundational Bayesian computation and frontier Bayesian theory. Presentation of Problem 6: optimal sampling strategies. Discussion and implementation of potential Bayesian experimental design methods for acquiring data from athletes. Presentation of Problem 7: using video data to compare planned and set play in team sports. Discussion and implementation of potential Bayesian spatio-temporal models. Communication of results.
• Day 5: Lectures on frontier Bayesian computation. Finalisation of problems 1-7. Extensions. Concluding remarks