[HAMA-809] Underflow in LogisticRegressionModel

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.6.1, 0.6.2
• Fix Version/s: 0.6.3
• Component/s: machine learning
• Assignee: Tommaso Teofili
• Reporter: Tommaso Teofili
• Created: 09/Oct/13 14:10
• Updated: 28/Nov/13 06:28
• Resolved: 28/Nov/13 06:28

Description

There's an underflow bug in LogisticRegressionModel when calculating the hypothesis output due to double limitations.
Therefore we should leverage BigDecimals for such operations as they allow for more precise calculation.
Also the log algorithm in LRM is wrong and needs to be fixed.