回答編集履歴
1
文章を整形
test
CHANGED
|
@@ -1,11 +1,17 @@
|
|
|
1
|
-
単純に、使える情報は多いほうが得だと思うので特に疑問は抱いていなかったのですが、
|
|
1
|
+
単純に、使える情報は多いほうが得だと思うので、特に疑問は抱いていなかったのですが、
|
|
2
2
|
|
|
3
|
-
考えてみると具体的にいつ役立つのか気になってきたのでせっかくなので「peephole connection paper」で検索して出てきた [こちらの論文](http://www.jmlr.org/papers/volume3/gers02a/gers02a.pdf) のAbstractを読むと
|
|
3
|
+
考えてみると具体的にいつ役立つのか気になってきたので、せっかくなので「peephole connection paper」で検索して出てきた [こちらの論文](http://www.jmlr.org/papers/volume3/gers02a/gers02a.pdf) のAbstractを読むと
|
|
4
4
|
|
|
5
|
-
We find that LSTM augmentedby “peephole connections” from its internal cells to its multiplicative gates can learn the fine distinction between sequences of spikes spaced either 50 or 49 time steps apart without the help of any short training exemplars.
|
|
5
|
+
> We find that LSTM augmentedby “peephole connections” from its internal cells to its multiplicative gates can learn the fine distinction between sequences of spikes spaced either 50 or 49 time steps apart without the help of any short training exemplars.
|
|
6
6
|
|
|
7
|
-
Without external resets or teacher forcing ,our LSTM variant also learns to generate stable streams of precisely timed spikes and other highly nonlinear periodic patterns.
|
|
8
7
|
|
|
8
|
+
|
|
9
|
+
> Without external resets or teacher forcing ,our LSTM variant also learns to generate stable streams of precisely timed spikes and other highly nonlinear periodic patterns.
|
|
10
|
+
|
|
11
|
+
|
|
12
|
+
|
|
9
|
-
This makes LSTM a promising approach for tasks that require the accurate measurement or generation of time intervals.
|
|
13
|
+
> This makes LSTM a promising approach for tasks that require the accurate measurement or generation of time intervals.
|
|
14
|
+
|
|
15
|
+
|
|
10
16
|
|
|
11
17
|
というので、時間の僅かな差を区別する時に役立つようですね。
|