Neural_Network
Perceptons(感知机)
input is 0 or 1, output is 0 or 1. + threshold
value(阈值) + weights(权重) + bias(偏置) : measure of how easy it is to
get the perceptron to output a 1
Activation Function
sigmoid neuron.

Just like a perceptron, the sigmoid neuron has
inputs, x1,x2,… . But instead of being just 0 or 1 , these inputs can
also take on any values between 0 and 1 . So, for instance, 0.638… is a
valid input for a sigmoid neuron. Output is between 0 and 1.
\[ \begin{eqnarray} \sigma(z) \equiv \frac{1}{1+e^{-z}}. \end{eqnarray} \]
Tanh

\[ \tanh (x)=\frac{2}{1+e^{-2 x}}-1 \]
ReLU

\[ \begin{eqnarray} \text{ReLU}(x) \equiv \max(0, x) \end{eqnarray} \]
COEN 6331
All articles in this blog are licensed under CC BY-NC-SA 4.0 unless stating additionally.
Comment