Dan Frankowski·Apr 14, 2022How to become a data scientist, part 3: tell people about your workIn part 1, I suggested you find a good problem. In part 2, I suggested you try to solve it. Here I discuss how to tell people what you did.
Dan Frankowski·Oct 7, 2021What if Recommendation Algorithms Like Facebook’s Grappled Directly with Bad Content?Maybe we can improve recommendations by giving our models a different goalA response icon1A response icon1
InTDS ArchivebyDan Frankowski·Aug 13, 2019Should you explain your predictions with SHAP or IG?Two different explanation algorithm types, best in different situations.A response icon1A response icon1
InTDS ArchivebyDan Frankowski·Jul 31, 2019Causality in model explanations and in the real worldYou can’t always change a human’s input to see the output.A response icon1A response icon1
InTDS ArchivebyDan Frankowski·Jul 22, 2019“Hey, what’s that?” Debugging predictions using explanationsWhat does debugging look like in the new world of machine learning models? One way uses model explanations.
InTDS ArchivebyDan Frankowski·Jun 3, 2019A gentle introduction to GA2Ms, a white box modelA gentle introduction to a white box machine learning model called a GA2M, a Generalized Additive Model (GAM) with interaction terms.A response icon1A response icon1
InTDS ArchivebyDan Frankowski·May 8, 2019Humans choose, AI does notFor the non-technical reader who sees scary headlines: every AI has a goal, usually a labeled dataset, precisely defined by a human.
InTDS ArchivebyDan Frankowski·Apr 23, 2019A gentle introduction to algorithmic fairnessA gentle introduction to issues of algorithmic fairness: some U.S. history, legal motivations, and four definitions with counterarguments.
InfiddlerlabsbyDan Frankowski·Mar 21, 2019Case study: explaining credit modeling predictions with SHAPAt Fiddler labs, we like explaining machine learning models. To learn more about how SHAP works, we applied it to predicting loan defaults.A response icon2A response icon2
Dan Frankowski·Feb 11, 2019Mary and John: using first name to predict sex in the US works quite wellSomeone’s first name is a good clue of their sex. Mary is probably female. John is probably male. How good a clue exactly?A response icon1A response icon1