GitHub user tillrohrmann opened a pull request:
https://github.com/apache/flink/pull/758
[ml] Rework of the optimization framework
This PR reworks the current optimization framework to make the
regularization part of the optimization algorithm. Furthermore, it consolidates
the `LossFunction` by defining a common interface for loss functions. The
`GenericLossFunction` allows to construct loss function by specifying the outer
loss function and the prediction function individually.
Additionally, this PR introduces some syntactic sugar which makes
programming with broadcast variables easier.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/tillrohrmann/flink mlSugar
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/flink/pull/758.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #758
----
commit e82782e5a165d9e3a348db5efe6c8188c9cbac72
Author: Till Rohrmann <trohrm...@apache.org>
Date: 2015-05-28T01:03:24Z
[ml] Adds syntactic sugar for map with single broadcast element. Rewrites
the optimization framework to to consolidate the loss function.
Adds closure cleaner to convenience functions on RichDataSet
Removing regularization from LossFunction and making it part of the
optimizer.
----
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
