Computers, Information, and Decision-Making
Peoples & Things
00:00
Is That Really an Improvement?
In a lot of computing, there are these tiny improvements. But testing for stotistical significance is not a common thing in computer science. Maybe you did a little better on this one data set, but does it work wisely on other data sets? Right? There are these sort of challenges people focus on. And focusing on those specific metrics about our accuracy can blinding you to all the other things that matter too.
Transcript
Play full episode
