On 7 April 2022, Professor Jianxin Pan, Dean of the Division of Science and Technology, UIC, gave a lecture on "Modelling and Learning Strategy for Medical/Health Data" to UIC students and faculty, introducing his recent research on data modelling in the field of medical and health.

Professor Jianxin Pan

He was a Professor of Statistics at the University of Manchester before joining UIC. He is a Fellow of the Royal Statistical Society, an Elected Member of the International Statistical Institute and a Turing Fellow of the Alan Turing Institute for Data Science and Artificial Intelligence. He has been serving as an Associate Editor for several prestigious academic journals, including Biometrics (2008-2018), Biostatistics and Epidemiology (2013-), Biometrical Journal (2016-), and Journal of Multivariate Analysis (2019-) and Electronic Journal of Statistics (2022-).

Before explaining the modelling case, Professor Pan introduced his research interests, mainly in mathematical statistics, applied statistics and data science. He then briefly described the idea behind precision medicine:a large number of variables but small sample sizes, which will be very challenging to traditional statistics methods, hence require statisticians to do research for new methods.

Finite mixture models

The first example was a classification of 51 countries in Europe based on average weekly mortality data. Professor Pan briefly introduced the finite mixture model and pointed out that in practical problems, we need to model both the mean and the variance. He talked about how this is about optimising a high-dimensional function in a high-dimensional space. Professor Pan then pointed out that this method could be used in many different areas, such as classifying the state of wages in Finland. The second example used data on joint pain in the elderly, collected every three years in the UK. Professor Pan pointed out that as the three sets of data were from the same person, correlations were taken into account and the three models were analysed simultaneously. The final conclusion is that women have a higher probability of having joint pain than men.

Q&A session

After the lecture, Professor Pan interacted with the audience and patiently answered the queries of teachers and students regarding the model, data and indicators.

Group photo of Professor Pan with students and teachers

Reporter: Qianqian Lu

Photographer: Qianqian Lu and Xiaojing Wei

Editors: Ms. Garbo Hu and Ms. Guoqiu Zhang