Add Gaussian Naive Bayes (rapidsai#4079)

This is a continuation of PR rapidsai#1763 and rapidsai#4053, to add Gaussian Naive Bayes. This is supposed to be merged after rapidsai#4053 Here is a comparison of cuML and SKLearn performance on Gaussian NB. This is done using a synthetic dataset generated by make_regression. The GPU used is a RTX 8000, and the CPU is i9-10920X @ 3.50GHz ![gaussian](https://user-images.githubusercontent.com/9810050/126572439-8982faa8-5ad1-4bca-91ab-76704050bf33.png) Linking issue rapidsai#1666 Authors: - Micka (https://github.com/lowener) - Corey J. Nolet (https://github.com/cjnolet) Approvers: - Corey J. Nolet (https://github.com/cjnolet) URL: rapidsai#4079
vimarsh6739 · Aug 9, 2021 · e8d4c40 · e8d4c40
1 parent 398ff4a
commit e8d4c40
Show file tree

Hide file tree

Showing 4 changed files with 532 additions and 0 deletions.
diff --git a/docs/source/api.rst b/docs/source/api.rst
@@ -267,6 +267,9 @@ Naive Bayes
 .. autoclass:: cuml.naive_bayes.BernoulliNB
     :members:
 
+.. autoclass:: cuml.naive_bayes.GaussianNB
+    :members:
+
 Stochastic Gradient Descent
 ---------------------------
 

diff --git a/python/cuml/naive_bayes/__init__.py b/python/cuml/naive_bayes/__init__.py
@@ -16,3 +16,4 @@
 
 from cuml.naive_bayes.naive_bayes import MultinomialNB
 from cuml.naive_bayes.naive_bayes import BernoulliNB
+from cuml.naive_bayes.naive_bayes import GaussianNB