Five Easy Things to fix Machine Learning
2019-12-28, 12:50–12:55, Borg

There are five simple rules to follow when researching in Machine Learning. They make your work easier to verify.


I have studied Computer Science and had to deal with many papers during my master's thesis. "Thanks" to poorly documented code and missing technical details I had a fair number of problems. As a consequence, I developed five easy to follow rules that improve the reproducability of any research significantly.

See also: Slides from the talk