I have trained a smaller neural network (4, 3, 1) with a smaller dataset (150 records, 4 features, 1 class). I trained the model with mini_batch_size = 1 and for 10 epochs. Now the trained model mapping all the test data to a fixed value. Could anyone explain why it is happening?
I was expecting to classify the dataset into true and false where 1/3 of the dataset is true.
