ignite-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From GitBox <...@apache.org>
Subject [GitHub] [ignite] dehasi commented on a change in pull request #6567: IGNITE-9978 Implement Compound Naive Bayes classifier
Date Tue, 04 Jun 2019 21:52:53 GMT
dehasi commented on a change in pull request #6567: IGNITE-9978 Implement Compound Naive Bayes
classifier
URL: https://github.com/apache/ignite/pull/6567#discussion_r290512101
 
 

 ##########
 File path: modules/ml/src/main/java/org/apache/ignite/ml/naivebayes/discrete/DiscreteNaiveBayesTrainer.java
 ##########
 @@ -49,6 +52,9 @@
     /** The threshold to convert a feature to a discrete value. */
     private double[][] bucketThresholds;
 
+    /** Feature ids which should be skipped. By defaut all features are processed. */
+    private Collection<Integer> featureIdsToSkip = emptyList();
+
     /** {@inheritDoc} */
     @Override public <K, V> DiscreteNaiveBayesModel fit(DatasetBuilder<K, V>
datasetBuilder,
                                                         Preprocessor<K, V> extractor)
{
 
 Review comment:
   @avplatonov That's a good idea. I played with `map` function in an extractor,   I can do
smth like `extractor = extractor.map(skipFeatures(featureIdsToSkip));` before call `fit`.
 (see my last commit in compound trainer) 
   But in this case, I make a copy of each vector.  Is it an acceptable approach?

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

Mime
View raw message