Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Probably typo in Chapter 7 Xgboost #1

Open
gelotran opened this issue Apr 17, 2020 · 4 comments
Open

Probably typo in Chapter 7 Xgboost #1

gelotran opened this issue Apr 17, 2020 · 4 comments

Comments

@gelotran
Copy link

Hi,

Can you check the formula in the part 7.4.3 Xgboost? I think there may be a typo there.

image

Then the formula 7.5 has to be adjusted accordingly.

Best,
Vu Tran

@shokru
Copy link
Owner

shokru commented Apr 17, 2020

Thanks for the feedback!
I've checked again, and it seems ok to me, given the 2 factor in front of the sum...

@gelotran
Copy link
Author

Hi,

Thank for feedback. I take a comparison with XGBoost website.

image

It seems that inside each tree leaf, we sum all the Hessian values but not a summation of \lamda.

Btw, your book is great for an R user working with Portfolio ML. I enjoy reading it a lot.

@shokru
Copy link
Owner

shokru commented Apr 17, 2020

True, but I prefer to take the simpler case when the loss is quadratic (easier to follow I find).
And for a quadratic loss, the Hessian is constant.
That's why I get differences w.r.t to the original derivation with "2" factors at some places.
Thanks for the feedback again: don't hesitate because we really want to continuously improve the book.

@gelotran
Copy link
Author

You're welcome. Thank for clarification

It's actually a good book. I hope that you can continue to work on it and looking to see further work from you.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants