Workshop presenters

Professor Yaser Samadi S. Yaser Samadi is an Associate Professor in School of Mathematical and Statistical Sciences and an affiliated faculty in School of Computing at Southern Illinois University Carbondale, IL, USA. He obtained his PhD in Statistics from the University of Georgia in 2014. He was a Research Fellow at SAMSI & Duke University in 2020-2021.

Title: Workshop on Symbolic Data Analysis

In our modern era of vast data availability, analyzing large datasets has become a routine challenge, given to advancements in technology. This workshop focuses on the analysis of symbolic data, which often arises when datasets are aggregated based on scientific criteria. Symbolic data includes various forms such as lists, intervals, and histograms, and finds applications across different fields including physical, medical, and social sciences. Additionally, certain datasets inherently possess symbolic values, such as species data, information with measurement uncertainties, financial data, and confidential data.

Symbolic data differs from classical data in that it represents distributions in multidimensional space rather than points. This workshop aims to shed light on the nature of symbolic data and their emergence in various contexts. By comparing methodologies for classical and symbolic data analysis, we highlight key differences through practical examples. Notably, we caution against the common practice of using classical surrogates, like aggregated means, which may lead to the loss of valuable information inherent in aggregated observation sets.

Professor Gary Sharp Gary Sharp is an academic in the Statistics Department at Nelson Mandela University, a Department he has had the honour to serve for more than a quarter of a century. During his time at NMU, he has served in many roles, including a seven-year tenure as Department Head. Gary’s research includes a wide range of topics, although his passion lies with analytics in sports, an area for which he is well known. Gary has supervised several post-graduate degrees, including several doctoral graduates. Gary is an active member of the South African Statistical Association (SASA) and has been instrumental in having the NMU Department of Statistics host annual SASA conferences. Gary is a Fellow and past President of SASA, an NRF rated researcher who has published nationally and internationally.

Title: Mathematical modelling in team selection sports

The ability to quantify an individual’s sport skill, provides analysts with the necessary quantitative tools to model the performances in a team setting. This workshop will present a mathematical modelling framework to select the optimal team or teams in scenarios where initially the best team seems intuitive, but rather quickly becomes more complex as the sporting code evolves. The examples used in this workshop originate in relay swimming, where a team consists of four swimmers, but are easily extended to other sporting codes. The focus in this workshop will be on formulating mathematical models and using available software to obtain optimal solutions.

Dr Janet van Niekerk Janet Van Niekerk, Ph.D. (Mathematical Statistics) is a statistician with experience in Bayesian modeling and computational statistics. She has published more than 25 articles in amongst others, Biostatistics, Statistical Methods in Medical Research, Biometrical Journal, Journal of Statistical Software and Computational Statistics and Data Analysis. She is an associate editor at Bayesian Analysis, and has presented various invited courses and workshops, most recently at the Centers for Disease Control and Prevention (CDC) in Atlanta, USA.

Title: Spatial modeling with INLA

Integrated Nested Laplace Approximations is a tool that provides accurate and efficient Bayesian inference of GAMM-type models. In this workshop I will show that most spatial models, being on a regular or irregular lattice, high resolution continuous data, or point processes can all be formulated as latent Gaussian models and thus a coherent methodology can be used to fit these models. The link between the Matern model and the weak solution of a specific SPDE is crucial for the approach in INLA and this will be briefly discussed. A new novel non-stationary Besag model for irregular lattice data will be presented. 

I will use the INLA package in R exclusively, and participants can work with me in the workshop if they have it installed beforehand. Link for the workshop: link