'Is it possible to train the neural network model with feature array of size (x,y,z) and label array of size (x,z)?

I am trying to train a Neural Network model which tries to predict the dynamic of a system in the next time system, providing current and some number of history information y as input.

number_samples= data_training.shape[0]- lookback - 1
X= np.zeros((number_samples,lookback+1,beta))
Y= np.zeros((number_samples,beta))
for i in range(number_samples):
   X[i] = data_training[i:i+lookback+1,2:]
   Y[i] = data_training[i+lookback+1,2:] - data_training[i+lookback,2:]

beta represents the dimension of my variable.

Is it possible to use the arrays X and Y to train my multi layer perceptron neural network ? and if yes should the data be shuffeled ? because I am interested in the dynamic of the system.

Thank you in advance

neural-network tf.keras

Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution	Source

'Is it possible to train the neural network model with feature array of size (x,y,z) and label array of size (x,z)?

Sources

Related Questions