Seonglae ChoExplained Variance Score
"X explains p% of the variance in Y" means that when we know X, the remaining variance of Y becomes (1-p) times the total variance
A metric that indicates how well a model explains the variance (variability) in the data
- 1.0: Model perfectly explains all variance
- 0.0: Model explains nothing (equivalent to always predicting the mean)
- Negative: Model makes variance worse (predictions are worse than mean)
FVU (Fraction of Variance Unexplained)
X explains Z% of the variance in Y — LessWrong
Recently, in a group chat with friends, someone posted this Lesswrong post and quoted: …
https://www.lesswrong.com/posts/E3nsbq2tiBv6GLqjB/x-explains-z-of-the-variance-in-y
