Thursday, April 26

AI&ML abstracts: review of possible effects of cognitive biases on interpretation of rule-based machine learning models

LinkedIn Google+ Pinterest Tumblr +

This paper investigates to what extent cognitive biases affect human understanding of interpretable machine learning models, in particular of rules discovered from data.

20 cognitive biases (illusions, effects) are covered, as are possibly effective debiasing techniques that can be adopted by designers of machine learning algorithms and software.

While there seems no universal approach for eliminating all the identified cognitive biases, it follows from analysis that the effect of most biases can be ameliorated by making rule-based models more concise.

Due to lack of previous research, our review transfers general results obtained in cognitive psychology to the domain of machine learning.

Read the full academic paper

Share.

About Author

Leave A Reply