[HAMA-791] Fix the problem that MultilayerPerceptron fails to learn a good hypothesis sometimes

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.6.3
• Labels: machine_learning
• Assignee: Yexi Jiang
• Reporter: Yexi Jiang
• Created: 07/Aug/13 03:25
• Updated: 07/Aug/13 23:35
• Resolved: 07/Aug/13 12:47

Description

Fix the problem that MultilayerPerceptron fails to learn a good hypothesis sometimes. If the weights of MLP initialized to very bad values, the model might fail the test.

Since the new model developed in HAMA-770 will replace the current old model, this problem will be simply fixed by increase the training iteration.