Logic is correct.
But I prefer to say experimental because the sample set is narrow. (300 columns)
6 inputs : Change , Low Band chg . , Mid Band chg ., Up Band chg . , change , histogram change.
1 output : Future bar change (Historical)
Training timeframe : 15 mins (Analysis TF > 4 hours (My opinion))
Learning cycles : 337
Training error: 0.009999
Input columns: 6
Output columns: 1
Excluded columns: 0
Training example rows: 301
Validating example rows: 0
Querying example rows: 0
Excluded example rows: 0
Duplicated example rows: 0
Input nodes connected: 6
Hidden layer 1 nodes: 8
Hidden layer 2 nodes: 0
Hidden layer 3 nodes: 0
Output nodes: 1
Learning rate : 0.6 Momentum : 0.8
More info :
EDIT : This code is open source under the MIT License. If you have any improvements or corrections to suggest, please send me a pull request via the github repository https://github.com/user-Noldo
On Line 188 : v = (-2.580743 * n_0 + -1.883627 * n_1 + -3.512462 * n_2 + -0.891063 * n_3 + -0.767728 * n_4 + -0.542699 * n_5 + 0.221093)
So we can create connection between layers and output . I suggest you to read article for more information : https://hackernoon.com/everything-you-need-to-know-about-neural-networks-8988c3ee4491
These are not big things as exaggerated, but basic things.
But if I did LSTM, I wouldn't share it :))
This command is very good but the more data, the more success it means.
300 columns too low.
If I get comprehensive indicator data, I can make more successful updates.
Then I plan to move to LSTM.
Thank you for feedback ! (y)