[SPARK-2841][MLlib] Documentation for feature transformations

[dbtsai]

Documentation for newly added feature transformations:

1. TF-IDF
2. StandardScaler
3. Normalizer

[SparkQA]

QA tests have started for PR 2068 at commit e339f64.

• This patch merges cleanly.

[SparkQA]

QA tests have finished for PR 2068 at commit e339f64.

• This patch passes unit tests.
• This patch merges cleanly.
• This patch adds the following public classes (experimental):
  • shift # Ignore main class (org.apache.spark.deploy.SparkSubmit) and use our own

[mengxr]

copy @atalwalkar

[atalwalkar]

This is too strong of a statement. Why not just say "Normalizing features to have unit variance and/or zero mean is very a common preprocessing step."

[dbtsai]

How about I say
"For example, RBF kernel of Support Vector Machines
or the L1 and L2 regularized linear models typically works better when all features have unit variance
and/or zero mean."

I actually have this statement from scikit documentation.
http://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html

[atalwalkar]

Your suggestion sounds good to me! Thanks.

[SparkQA]

QA tests have started for PR 2068 at commit 0a8fd34.

• This patch does not merge cleanly!

[SparkQA]

QA tests have finished for PR 2068 at commit 0a8fd34.

• This patch fails unit tests.
• This patch does not merge cleanly!

[dbtsai]

@atalwalkar and @mengxr I just addressed the merge conflict. I think it's ready to merge. Thanks.

[SparkQA]

QA tests have started for PR 2068 at commit 109f324.

• This patch merges cleanly.

[SparkQA]

Tests timed out after a configured wait of 120m.

[mengxr]

LGTM. Merged into master and branch-1.1! Thanks for helping on the documentation!!