Loss functions for daal4py GBT model and other models

[tcs_hormazd]

Hello,

Where can we find documentation about which specific loss functions we can pass to daal4py models (such as the  Gradient Boosted Regression model)?

[JaideepK_Intel · Accepted solution]

Hi,

Thank you for posting in Intel Communities.

For Regression Gradient Boosted Trees the loss function is squared and for Classification Gradient Boosted Trees loss function is crossEntropy. Please go through below documentation:

1. https://oneapi-src.github.io/oneDAL/daal/algorithms/gradient_boosted_trees/gradient-boosted-trees-regression.html
2. https://oneapi-src.github.io/oneDAL/daal/algorithms/gradient_boosted_trees/gradient-boosted-trees-classification.html
3. https://intelpython.github.io/daal4py/algorithms.html#gradient-boosted-regression
4. https://github.com/intel/scikit-learn-intelex/tree/master/examples/daal4py

Thanks,

Jaideep

[tcs_hormazd]

Thank you, Jaideep. Does the Regression Gradient Boosted Tree accept any loss function for the loss function parameter other than the default 'squared' loss?

[JaideepK_Intel]

Hi,

We recommend you to use scikit-learn-intelex package instead of daal4py.

Intel(R) Extension for Scikit-learn contains scikit-learn patching functionality that was originally available in daal4py package.

please go through below link:

https://github.com/intel/scikit-learn-intelex

daal4py Gradient Bossting Classification model creation from XGBoost example:

https://github.com/intel/scikit-learn-intelex/blob/master/examples/daal4py/gbt_cls_model_create_from_xgboost_batch.py

sorry to inform that in daal4py for Regression Gradient Boosted Trees model and for Classification Gradient Boosted Trees models support only one loss function i.e squared and crossEntropy.

Thanks,

Jaideep

[JaideepK_Intel]

Hi,

If this resolves your issue, make sure to accept this as a solution. This would help others with similar issue. Thank you!

Thanks,

Jaideep

[tcs_hormazd]

Hi Jaideep,

Thank you for your help with my query. Sorry for the delay. I have accepted one of the replies as a solution.

Best,

Hormazd

[JaideepK_Intel]

Hi,

Glad to know that your issue is resolved. If you need any additional information, please post a new question as this thread will no longer be monitored by Intel.

Have a great day ahead.

Thanks,

Jaideep