[jira] [Created] (IGNITE-9145) [ML] Add different strategies to index labels in StringEncoderTrainer

[Anton Vinogradov (Jira)]

Aleksey Zinoviev created IGNITE-9145:
----------------------------------------

             Summary: [ML] Add different strategies to index labels in StringEncoderTrainer
                 Key: IGNITE-9145
                 URL: https://issues.apache.org/jira/browse/IGNITE-9145
             Project: Ignite
          Issue Type: Improvement
          Components: ml
            Reporter: Aleksey Zinoviev
            Assignee: Aleksey Zinoviev
             Fix For: 2.7


The main idea to add a few strategies of indexing: sorting and so on.

Currently it supports only one strategy (most popular with zero and less popular with the max index size).

There are can be a few options
 * 'frequencyDesc': descending order by label frequency (most frequent label assigned 0)
 * 'frequencyAsc': ascending order by label frequency (least frequent label assigned 0)
 * 'alphabetDesc': descending alphabetical order
 * 'alphabetAsc': ascending alphabetical order

 

Please, update the method **transformFrequenciesToEncodingValues and add the strategy as a parameter of trainer.

 



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)