Louloumonkey
Louloumonkey

Reputation: 69

XGBoost and count:poisson

The XGBoost library lists count:poisson as one of its learning task parameters. The official documentation is pretty limited and only says:

Poisson regression for count data, output mean of Poisson distribution

Does anybody know a bit more about how this is calculated, as the above is quite confusing?

Alternatively, I've been trying to locate in the XGBoost GitHub repo the source code where count:poisson is defined, but to no avail. Does anybody know where I can locate this please? Aka, the specific file that details how this is calculated?

Upvotes: 1

Views: 307

Answers (0)

Related Questions