[SPARK-13089][ML] [Doc] spark.ml Naive Bayes user guide and examples #11015
Conversation
|
Test build #50531 has finished for PR 11015 at commit
|
hhbyyh
changed the title
| package org.apache.spark.examples.ml; | ||
|
|
||
| // $example on$ | ||
| import org.apache.spark.SparkConf; |
There was a problem hiding this comment.
I would not include SparkConf or JavaSparkContext
|
test this please |
|
Test build #55389 has finished for PR 11015 at commit
|
|
Test build #55471 has finished for PR 11015 at commit
|
|
Thanks for the review. Updated according to the comments. |
| probabilistic classifiers based on applying Bayes' theorem with strong (naive) independence | ||
| assumptions between the features. More information about the spark.ml implementation can be | ||
| found further in the section on [Naive Bayes in MLlib](mllib-naive-bayes.html#naive-bayes-sparkmllib). | ||
|
|
There was a problem hiding this comment.
I think it's better to clarify ml.NaiveBayes supports Multinomial NB and Bernoulli NB. Meanwhile, we should provide the link to corresponding documents. You can refer the NaiveBayes API doc.
There was a problem hiding this comment.
Thanks for taking a look. The wiki link already provided a good overall introduction to Naive Bayes. I'll add some clarification. And in the mllib documents, it clarifies naive Bayes supports both Multinomial and Bernoulli.
|
Test build #55596 has finished for PR 11015 at commit
|
| .setLabelCol("label") | ||
| .setPredictionCol("prediction") | ||
| .setMetricName("precision") | ||
| val accuracy = evaluator.evaluate(predictions) |
There was a problem hiding this comment.
I would call it "precision" instead of "accuracy" since it might confuse people (even though they are the same here).
|
Just that 1 comment |
|
Test build #55694 has finished for PR 11015 at commit
|
|
LGTM |
|
Thanks @jkbradley |
4 participants
jira: https://issues.apache.org/jira/browse/SPARK-13089
Add section in ml-classification.md for NaiveBayes DataFrame-based API, plus example code (using include_example to clip code from examples/ folder files).