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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14326961#comment-14326961 ] Jesse Daniels commented on MAHOUT-1089: --- I haven't studied the code in depth but I

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14322615#comment-14322615 ] clem clem commented on MAHOUT-1089: --- Hi, I have been using the SGD matrix factorizatio

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13695752#comment-13695752 ] Sebastian Schelter commented on MAHOUT-1089: Peng, please open a new ticket f

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13695745#comment-13695745 ] Peng Cheng commented on MAHOUT-1089: Code is slick! But apparently there is no multi-

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13486445#comment-13486445 ] Hudson commented on MAHOUT-1089: Integrated in Mahout-Quality #1723 (See [https://builds

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13486355#comment-13486355 ] Zeno Gantner commented on MAHOUT-1089: -- Another general comment: to my experience, t

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13486316#comment-13486316 ] Zeno Gantner commented on MAHOUT-1089: -- With learning rate decay 0.99: 0.075 5 0.01

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13486309#comment-13486309 ] Zeno Gantner commented on MAHOUT-1089: -- w/o learning rate decay, good parameter comb

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13483049#comment-13483049 ] Zeno Gantner commented on MAHOUT-1089: -- I plan to add a learning rate decay feature,

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13466673#comment-13466673 ] Ted Dunning commented on MAHOUT-1089: - Shuffling in each iteration is good practice,

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

[ https://issues.apache.org/jira/browse/MAHOUT-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13466659#comment-13466659 ] Sebastian Schelter commented on MAHOUT-1089: Hi Zeno, great work! I'm polish