Skip to main content
Thomas Wagenaar's user avatar
Thomas Wagenaar's user avatar
Thomas Wagenaar's user avatar
Thomas Wagenaar
  • Member for 7 years, 5 months
  • Last seen more than a month ago
6 votes
1 answer
534 views

Recurrent networks mimicking previous / current input

5 votes
1 answer
2k views

Is it better to split sequences into overlapping or non-overlapping training samples?

5 votes
1 answer
6k views

Neural network failing because of Infinity

3 votes
2 answers
1k views

Dropout effectiveness on small neural networks

3 votes
2 answers
487 views

Is the NEAT algorithm unbalanced?

3 votes
2 answers
1k views

Why is the default learning rate for Adadelta so low in Keras?

3 votes
2 answers
2k views

Proper shape of LSTM dataset for keras

3 votes
2 answers
862 views

Can a convolutional NN be made with perceptrons?

3 votes
0 answers
348 views

Should I use random fixed mutation or random percentual mutation?

2 votes
1 answer
664 views

Normalization - which is better?

2 votes
1 answer
281 views

Do different species create offspring (NEAT)?

1 vote
1 answer
96 views

Is nested mutation effective?

1 vote
1 answer
420 views

Are momentum and (mini-) batch training compatible? [closed]

1 vote
1 answer
32 views

Testing set 40x as big as training set?

1 vote
0 answers
442 views

When to stop episode during reinforcement learning

1 vote
0 answers
52 views

Can we derive the reward function from the behaviour of a trained agent?

1 vote
1 answer
17 views

Should identical inputs always have shared weights to the next layer?

1 vote
0 answers
117 views

Forcing the mean of predictions to converge to true value

1 vote
0 answers
106 views

Algorithm for neuron activation order