Seminar: Cognitive biases and interpretability of inductively learnt rules

This talk will report on investigation of  the effect of cognitive biases on human understanding of machine learning models, focusing on inductively learnt rules. We will cover applicable psychological phenomena, possible debiasing techniques, as well considerations for “interpretability bias”, which can be embedded into learning algorithms. We will also discuss the relation between inductive biases (machine learning) and cognitive biases (psychology).

Some work-in-progress related to the talk is at https://sites.google.com/site/kliegr/research/cognitive-biases-and-interpretability-of-machine-learning-models.