Explainable predictions: an introduction to model building within the supervised learning framework.
- Date: –15:00
- Location: Biomedicinskt centrum, BMC C4:305
- Lecturer: Marcin Kierczak, PhD
- Organiser: Department of Cell and Molecular Biology
- Contact person: Marcin Kierczak
The Department of Cell and Molecular Biology hereby invites all interested to a docent (associate professorship) lecture in subject Bioinformatics.
Experiments in life sciences generate larger and larger quantities of data and these data have to be analysed in an efficient and reliable way in order to give us insights into the nature of biological processes. The so-called supervised learning framework is a common approach to modelling and it relies on using available, labelled data to build models capable of predicting outcomes for previously unseen data points. In this lecture, I will give an overview of a typical modelling process: from input data engineering to building and evaluating predictive models using various machine learning techniques, including some deep learning methods. We will also look at various properties of selected modelling techniques, e.g. interpretability and explainability of the models.
The lecture is an obligatory teaching test for those applying for admittance as docent (associate professor) and it should be possible for students and others with basic academic education in the relevant field to follow it. The lecture will last for 45 minutes and afterwards the audience may ask questions. The lecture will be given in English.
Chairperson: Professor Jan Komorowski
Representative of the Associate Professorship Board: Professor Monika Schmitz