Center for Basic Machine Learning Research in Life Science
Machine learning (deep learning) has over the past decade led to a series of major breakthroughs in domains such as image analysis, computer vision, self-driving cars, natural language processing, chatbots, etc. The underlying methods development has been pioneered by both academic groups and technology companies such as Google, Facebook, Amazon, Microsoft, and Tesla. Applications in life science, however, have not yet seen the same level of attention, meaning that the methods in this domain are generally less efficient, less developed, and less explainable.
The Center for Basic Machine Learning Research in Life Science focuses on development of fundamental machine learning algorithms and methods tailored to life science applications, such as protein engineering and optimization, sequence variation, genomics, medical imaging, drug discovery, etc. These fields are characterized by:
- the data being heterogeneous, noisy, and sometimes even ambiguous;
- a need for ‘explainable AI’ (as opposed to black-box models) to build confidence and ensure real world implementation of the methods, particularly in the medical and clinical domain;
- a strong desire to go beyond simple prediction to uncover mechanistic insight from the models (e.g., to inspire research to develop new treatments of disease);
- a need for models which can mix unbiased learning from data with incorporation of existing domain knowledge.
The new center gathers some of the leading machine learning researchers in Denmark and will contribute to both teaching and research. It will foster a new generation of excellent methodology-oriented data scientists with enough knowledge in life science to advance real world practice.
The center will train PhD students and post-docs, organize international summer schools and workshops in collaboration with the elite academic European machine learning network ELLIS and offer at least two new MSc courses within the topics of the center.