Explain how balance_classes is triggered

[exalate-issue-sync]

Add into documentation how sampling is done for H2O3.

Suggestion to http://docs.h2o.ai/h2o/latest-stable/h2o-docs/data-science/algo-params/balance_classes.html:

With `balance_classes` enabled, it uses either `class_sampling_factors` or `max_after_balance_size` to control the sampling. 
* `class_sampling_factors` takes a list of numbers which would be the sampling rate for each class. A value of '1' would not change the sample rate for a class, but setting it to '0.5' would reduce its sampling by half and '2' would double its sample rate. 
* Alternatively, you can utilize `max_after_balance_size` which is the max relative size your training data could be grown. By default it is 5; this will oversample the data to rebalance the training data. The max it can grow to is 5x larger than your original data, hence, the value of 5. If you have many rows and prefer to under-sample the majority class, you can set max_after_balance_size to be < 1

[h2o-ops]

JIRA Issue Migration Info

Jira Issue: PUBDEV-7699
Assignee: hannah.tillman
Reporter: Neema Mashayekhi
State: Resolved
Fix Version: 3.30.1.1
Attachments: N/A
Development PRs: Available

Linked PRs from JIRA

#4815