[FLINK-1979] Implement Loss Functions

• Type: Improvement
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Component/s: Library / Machine Learning
• Labels: ML pull-request-available
• Assignee: Soila Kavulya
• Reporter: Johannes Günther
• Created: 06/May/15 12:08
• Updated: 28/Feb/19 22:48
• Resolved: 12/Feb/17 19:42

Description

For convex optimization problems, optimizer methods like SGD rely on a pluggable implementation of a loss function and its first derivative.

Issue Links

• is part of: FLINK-2013 Create generalized linear model framework
• relates to: FLINK-1889 Create optimization framework
• relates to: FLINK-1807 Stochastic gradient descent optimizer for ML library
• links to: GitHub Pull Request #656
• links to: GitHub Pull Request #656
• links to: GitHub Pull Request #1985
• links to: GitHub Pull Request #1985