Chao Xing - 版本历史

2015年10月26日 (一) 06:50 Xingchao

2015-10-26T06:50:50Z

2015年10月7日 (三) 09:05 Xingchao

2015-10-07T09:05:23Z

2015年10月7日 (三) 09:05 Xingchao

2015-10-07T09:05:05Z

Xingchao：以“Last Week: Last Week Solution: 1. DNN program has some problems, fix these & DNN contain same performance to Linear. Problem is training process, training sa...”为内容创建页面

2015-10-07T09:04:12Z

以“Last Week: Last Week Solution: 1. DNN program has some problems, fix these & DNN contain same performance to Linear. Problem is training process, training sa...”为内容创建页面

新页面

Last Week:
Last Week Solution:
1. DNN program has some problems, fix these & DNN contain same performance to Linear.
Problem is training process, training samples are equal to speakers' so this will lead model less compute 50% training samples.
Plan to do:
(Hold)
Problem:
1. DNN still not better than linear model.

This Week:
Solution:
1. DNN with dropout method, and add SAA to reduce randomly properties.
Plan to do:
1. Run for a small vocabulary size, such as 5000.
2. Run and test some ideas.

@@ 第1行： / 第1行： @@
 Last Week:
 Plan to do:
-This Week:
+. Learning RNN, LSTM, Memory.
-  Solution:
+. Writing technique report about Matrix Factorization.
-. DNN with dropout method, and add SAA to reduce randomly properties.
+. Draw deep into different theory of matrix decompose.
-  Plan to do:
-. Run for a small vocabulary size, such as 5000.
-. Run and test some ideas.

@@ 第1行： / 第1行： @@
 Last Week:
 Last Week Solution:
-. DNN program has some problems, fix these & DNN contain same performance to Linear.
+   (Hold)
-     Problem is training process, training samples are equal to speakers' so this will lead model less compute 50% training samples.
 Plan to do:
    (Hold)
 Problem:
-. DNN still not better than linear model.
+   (Hold)
-−