Skip to main content
Share Your Experience: Take the 2024 Developer Survey
10 votes
Accepted

When to use Tanh?

Using tanh in hidden layers require careful initialization of network weights and works best with the input features normalized within the same range as output (i.e. -1 to 1). It have expensive ...
Leo's user avatar
  • 425
2 votes

Neural network for specific numbers from a range (Q learning)

def action(self, state): if np.random.rand() > self.epsilon: return np.random.randint(0,4) return np.argmax(self.model.predict(state)) You do ...
foreverska's user avatar
  • 1,263
2 votes
Accepted

Neural network for specific numbers from a range (Q learning)

I am not able to understand how can I tell the neural network that it can take only one of the following values You don't have to. The neural network in Deep Q Learning (the DQN) is not configured to ...
Neil Slater's user avatar
  • 32.4k
2 votes

Why are the tutorials and built-in datasets giving us examples that simply do not work?

A tutorial is not keys to a production grade system. They are just to get one familiar with the basic concepts. To move toward a system that works on more than the provided examples, takes quite a ...
foreverska's user avatar
  • 1,263
2 votes

Why do we need complex neural network designs?

So: "even a powerful neural network like XGBoost": XGBoost is not a neural network "we can even train a simple MLP neural network with higher accuracy": higher with respect to ...
Alberto's user avatar
  • 2,153
2 votes

Sparse Cross Entropy

So class 0 has loss $− 0 \log \hat y=0$ and class 9 has loss $−9 \log \hat y$ ? This is clearly not correct. The correct implementation takes as input a vector $\hat y$ with 10 elements and you want ...
Sycorax's user avatar
  • 473
1 vote

How do neural-network chess engines select moves?

That's up to you and mainly depends on your budget: if you want it to be as cheap as possible, you probably want to forward the state on the policy network and pick the move with the highest ...
Alberto's user avatar
  • 2,153
1 vote

Can anyone please explain the Recurrent Neural Network calculation shown in the picture?

This is the RNN architecture. Image is taken from https://en.wikipedia.org/wiki/Recurrent_neural_network. Here $U = B$, $V = A$, and $W = C$. We are using ReLU as the activation function. $\phi(x) = ...
vishal's user avatar
  • 11

Only top scored, non community-wiki answers of a minimum length are eligible