#ML_Wednesday 
Today is a commemoration day, but it doesn't mean that I can take time to read some ML!
Return to the basics for today: the Bias-Variance Tradeoff!

Today is a commemoration day, but it doesn't mean that I can take time to read some ML!
Return to the basics for today: the Bias-Variance Tradeoff!
To begin with, here is an interesting math-free introduction to the topic: https://elitedatascience.com/bias-variance-tradeoff
It explains what happens, but not really WHY it happens...
It explains what happens, but not really WHY it happens...
With a little bit more maths to explain the theory behind, another article, by Seema Singh: https://towardsdatascience.com/understanding-the-bias-variance-tradeoff-165e6942b229
(the graphics are around the same)
(the graphics are around the same)
Now something very cool I didn't know about... double descent in the test error curve! 

And a very good thread by @daniela_witten about it: https://twitter.com/daniela_witten/status/1292293102103748609
It's very clear and pedagogic so I'll have nothing to add about this topic!


And a very good thread by @daniela_witten about it: https://twitter.com/daniela_witten/status/1292293102103748609
It's very clear and pedagogic so I'll have nothing to add about this topic!