Now that we've made our #introductions, let's start to get some science into our feeds! For #MLMonday, I want to share Cynthia Rudin's excellent talk on explainable #ML "Do Simpler Machine Learning Models Exist and How Can We Find Them?" in the NSF CISE Distinguished Lecture Series. Check out the NSF page https://beta.nsf.gov/events/do-simpler-machine-learning-models-exist-how-can or go straight to the video https://players.brightcove.net/679256133001/NkgrDczuol_default/index.html?videoId=6314127123112
@aliceschwarze This talk was really great, while I was in a PhD program I was working on Reconstructability Analysis, which searches for the simplest models in a set of possible models, and also produces interpret-able decision trees.
Definitely agree that explain-ability and interpret-ability are extremely important for high stakes automated decision making.
Very interested in Rashomon sets and how one determines the size of the Rashomon set of models compared to the set of all models?
