On February 23, 2023, we conducted a prediction modelling workshop in the hybrid format. This was a half-day intensive workshop to learn about how to conduct research to develop prediction models in healthcare. The lecturer was Dr. Orestis Efthimiou at the Institute of Primary Health Care, University of Bern, Switzerland. We invited him as Kyoto University Visiting Scholar for two weeks in Kyoto University School of Public Health.

In this workshop, the participating students and faculty members learned about understanding that there are different types of prognostic research using various prediction modelling techniques. Dr. Efthimiou taught us such contents as basic statistical models, methods, and concepts used in prediction modelling.

The participants learned the way of thinking regarding how to deal with overfitting and optimism, as well as dealing with missing data.

In the last part of the lecture, Dr. Efthimiou introduced various methods of machine learning and use of AI in building prediction models. The lecture included methods of supervised machine learning, such as artificial neural networks (“deep learning”), support vector machines, decision trees, random forest, boosting and other techniques.