Copy

Newsletter

Hey there, you just received the monthly depends-on-the-definition Newsletter for December 2019. The year is coming to an end and I hope it was a happy and successful one for you. And you learned a lot of new things.

Either way, I wish you a very happy New Year 2020!

Post of the month:
How explainable AI fails and what to do about it
Nowadays, there are a lot of people talking about and advertising the methods of “Explainable AI”. The derived explanations are often not reliable, and can be misleading.
We will discuss the problems with Explainable ML and show how some of those problems can be solved with interpretable machine learning.
Read More

Paper pick

Cynthia Rudin: "Stop Explaining Black Box Machine Learning Models for High Stakes Decisions and Use Interpretable Models Instead"

Tips & Tricks

The python package "Scikit-Lego" has some very useful support to deal with fairness issues in machine learning.

https://scikit-lego.readthedocs.io/en/latest/fairness.html

Recommended reading

In the gradient article "An Epidemic of AI Misinformation" the author Gary Marcus warns about the risks and dangers of overpromising and over-hyping in the field of AI. He also recommends what to do instead.
Facebook
Twitter
Link
Website
Copyright © 2019 depends-on-the-definition, All rights reserved.