Invited Speakers

Ameen Abu-Hanna, University of Amsterdam

Prediction models: opportunities and limitations

Abstract: In this talk I will address opportunities and limitations of biomedical prediction models for their perceived use and describe analytic challenges facing developers of such models from messy observational data. In particular, I will address the following topics: The use of prediction models in decision support and benchmarking; The dilemma of developing aetiological or prediction models; Enriching models with novel predictors and how to assess their added value; Novel temporal and external validation frameworks; and reporting standards you need to know about when working with messy observational data. These topics will be illustrated by real-world applications.

Bio sketch: Ameen Abu-Hanna is professor of Medical Informatics and head of the department of Medical Informatics at the Academic Medical Center at the University of Amsterdam. His academic background is in Computer Engineering, Computer Science, and AI; and his research interests include statistical machine learning and clinical decision support. He is a former associate editor of Journal of Biomedical Informatics and former president of the European Society of AI in Medicine. In 2017, he was elected a founding member of the International Academy of Health Sciences Informatics.

Blaz Zupan, University of Ljubljana

Data Science for Everyone

Machine learning methods are drivers of change in health care. They will become omnipresent. Yet, few health care professionals understand the basics of data science, and even fewer can build models using their own data. In this talk, I will explain how anybody who can spare a few hours can learn the mechanisms behind data mining. The training will give people enough intuition to benefit from their data and to recognize opportunities that they had not seen. In this lecture, I will walk through selected cases we use to explain data mining to a lay audience. The training explains concepts from machine learning that include clustering, classification, data visualization, and deep model-based embedding.

Bio sketch: Blaz Zupan is a Professor of Computer Science at University of Ljubljana, where he heads the bioinformatics lab. His research revolves around techniques for data fusion and data visualization. His lab developed Orange, an evolving data mining suite with a visual programming environment. He also enjoys writing the scripts for YouTube videos that explain Orange and data science. Blaz and his colleagues enjoy preparing courses that introduce data science to diverse audiences in biomedicine, pharmaceuticals, the humanities and commerce.