search

zackchase / mxnet-the-straight-dope

An interactive book on deep learning. Much easy, so MXNet. Wow. [Straight Dope is growing up] ---> Much of this content has been incorporated into the new Dive into Deep Learning Book available at https://d2l.ai/.
https://d2l.ai/
Apache License 2.0
2.56k stars 725 forks source link

Mistake in "Linear regression from scratch" #313

Open mseeger opened 7 years ago

mseeger commented 7 years ago

In this chapter, you say the noise on the real function is drawn from a Gaussian with mean 0 and variance 0.1. However, your code draws the noise from a Gaussian with mean 0 and variance 0.0001, which is a lot less. You multiply a N(0,1) Gaussian by 0.01 (the stddev). To get variance 0.1, you'd have to multiply by sqrt(0.1).

zackchase commented 7 years ago

Thanks for the heads up! There have been some code changes need to revise to make code and text agree (of coyrse, maybe it was borked in the first place!)

Please let us know when you see other issues!