How to mask a loss function (mae) in Keras?-CodePudding

I am trying to implement a custom loss function for Keras LSTM, which would represent mask_MAE.

def mask_MAE (y_true, y_pred, mask):# mask = 0 or 1
    mae = K.abs(y_pred - y_true) * mask
    return K.sum(mae)/K.sum(mask)

CodePudding user response：

I found an answer to my question. I am working with LSTM and 80 is num_steps

def GBVPP_loss(y_true, y_pred, cols = 80):
   u_out = y_true[:, cols: ]
   y = y_true[:, :cols ]
   w = 1 - u_out
   mae = w * tf.abs(y - y_pred)
   return tf.reduce_sum(mae, axis=-1) / tf.reduce_sum(w, axis=-1)
...
history = model.fit(X_train, np.append(y_train, u_out_train, axis =1), 
                 validation_data=(X_valid, np.append(y_valid, u_out_valid, axis =1)), 
                 epochs=EPOCH, batch_size=BATCH_SIZE,
                 verbose=0,
                 callbacks=[lr])

CodePudding user response：

A custom keras loss function can have only two parameters- y_true & y_pred.
So, you cannot use that mask parameter as you have done in your code.