0
$\begingroup$

I am able to fit this autoencoder to my sequence in order to reconstruct it.

However, how would I be able to walk this forward 3 timesteps to get [[11.0], [12.0], [13.0]]?

import numpy as np
from keras.models import Sequential
from keras.layers import Dense, Flatten, LSTM


# --- Data Prep ---
arr_train = [[[1.0], [2.0], [3.0], [4.0], [5.0], [6.0], [7.0], [8.0], [9.0], [10.0]]]
arr_train = np.array(arr)
# arr_train.shape == (1, 10, 1): [samples, timesteps, features]

arr_test = arr_train # test the reconstuction.


# --- Model Design ---
timesteps = arr_train.shape[1]
features = arr_train.shape[2]
outputs = arr_train.shape[1]#10

model = Sequential()
model.add(LSTM(60, input_shape=(timesteps, features)))
model.add(Dense(outputs))
model.compile(loss='mse', optimizer='adamax')

model.fit(
    arr_train
    , arr_test
    , epochs=200
    , batch_size=1
)
# loss == 0.01

If I try to set output=13 it says it must be 10.

$\endgroup$
0
$\begingroup$

Did you try this?

Xnew = [[...], [...]]
ynew = model.predict_classes(Xnew)
$\endgroup$
1
  • $\begingroup$ Hi Nicolas. It is not a classification problem. $\endgroup$ – HashRocketSyntax Jun 18 at 11:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.