[MAHOUT-1089] SGD matrix factorization for rating prediction with user and item biases

• Type: New Feature
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.7
• Fix Version/s: 0.8
• Assignee: Sebastian Schelter
• Reporter: Zeno Gantner
• Created: 30/Sep/12 21:01
• Updated: 31/Jan/24 22:11
• Resolved: 29/Oct/12 20:52

Description

A matrix factorization that is trained with standard SGD on all features at the same time, in contrast to ExpectationMaximizationFactorizer, which learns feature by feature.

Additionally to the free features it models a rating bias for each user and item.