big data

learn_bAIome offers workshops and trainings in biomedical AI/data science with tailored formats that take into account background, programming skills and intensity to provide unique, focused, and effective courses. These courses are free and open to students, clinicians, and researchers across academic institutions in Hamburg.

Lecturer: Dr. Lorenz Adlung, I. Department of Medicine, Hamburg Center for Translational Immunology (HCTI), bAIome, UKE

Prerequisites: Intrinsic motivation to learn about infection and inflammation using your computer.

Description: This workshop is open to all students, researchers and clinicians who want to learn how we use (“big”?) data and computational modelling for discovery and rational intervention in infection and inflammation. In today’s biomedical research, the bottleneck has shifted, and for the first time, data generation is no longer the rate-limiting step in scientific progress, but rather: data analysis. We will discuss current trends and show how we can use mathematical concepts and analytical thinking to address unmet clinical needs in influenza infection and inflammatory bowel disease. The workshop will be in presence and therefore each participant should bring their own laptop or ipad.

Topics

• “big” data and code repositories
• mathematical concepts relevant to infection and inflammation
• computational modelling of murine influenza infection
• identification of treatment responders in inflammatory bowel disease
• discussion of the future of bAIomedical research

Techniques of machine learning (ML) find a rapidly increasing range of applications touching upon social, economic, and technological aspects of everyday life. They are also being used with great enthusiasm to fill in gaps in our scientific knowledge by data-based modelling approaches. I have followed these developments for a while with interest, concern, and mounting disappointment. When these technologies are employed to take over decisive functionality in safety-critical applications, we would like to exactly know how to guarantee their compliance with pre-defined guardrails and limitations. Moreover, when they are utilized as building blocks in scientific research, it would violate scientific standards -in my opinion- if these building blocks were used without a throrough understanding of their functionality, including inaccuracies, uncertainties, and other pitfalls. In this context, I will juxtapose (a subset of) deep neural network methods with the family of entropy-optimal Sparse Probabilistic Approximation (sSPA) techniques developed recently by Illia Horenko (RPTU Kaiserslautern-Landau) and colleagues.