Htykuut

I have 1D signal data which I am provided, its basically synthetic data, not realistic.

I have 1000 CSV files for the same which have the following variables.

1=depth(index), 2=signal value(1D signal), 3=event id, 4=event id as a class (target variable), 5=event and class (not useful)

0           ,          3            ,       0         ,             0                    ,                0

1           ,          5            ,       0         ,             0                    ,                0

2           ,          2            ,       0         ,             0                    ,                0

3           ,          3            ,       1         ,             1                    ,                1000

4           ,          6            ,       0         ,             1                    ,                1

5           ,          1            ,       0         ,             1                    ,                1

6           ,          2            ,       2         ,             2                    ,                2000

7           ,          6            ,       0         ,             2                    ,                2

8           ,          2            ,       0         ,             2                    ,                2

9           ,          6            ,       0         ,             2                    ,                2

I want to train multiple classification models using keras, pytorch, seq2seq.

I haveing various idea in my mind how to create the data so it looks better and feed them in my TCN/CNN networks.

feeding as a 1D array as feature and 1D array as label in a classification model does not make a good sense.

I need some ideas, if someone has done it before.

Thanks in Advance.