pyml.neural_network.layer.activation.softmax.Softmax#
- class Softmax[source]#
Bases:
_ActivationSoftmax activation function
The softmax activation function is defined as \(\sigma (\mathbf {z} )_{j}={\frac {e^{z_{j}}}{\sum _{k=1}^{K}e^{z_{k}}}}\).
Methods
__init__Computes the backward step
Computes a forward pass
Converts outputs to predictions
Set adjacent layers which are needed for the model to iterate through the layers.
- backward(dvalues)[source]#
Computes the backward step
The derivative of the softmax function will be calculated as follows: \(\frac {\partial S(z_{i})}{\partial z_{j}} = {\begin{cases} S(z_{i}) \cdot (1 - S(z_{i})) & \text{if } i = j \\ - S(z_{i}) \cdot S(z_{j}) & \text{if } i \neq j \end{cases}}\).
- Return type:
- Parameters:
dvalues (numpy.ndarray) – Derived gradient from the previous layers (reversed order).
- forward(inputs)[source]#
Computes a forward pass
The softmax activation function is defined as \(\sigma (\mathbf {z} )_{j}={\frac {e^{z_{j}}}{\sum _{k=1}^{K}e^{z_{k}}}}\).
- Return type:
- Parameters:
inputs (np.ndarray) – Input values from previous neural layer.
- predictions(outputs)[source]#
Converts outputs to predictions
The softmax activation function calculates the confidence for each class. Since the the softmax function does not only support binary tasks, but also multiclass problems, the user must receive the index of the class with the highest confidence score.
- Return type:
ndarray- Parameters:
outputs (numpy.ndarray) – Output computed by the softmax activation function
- Returns:
Matrix containing the class indices with the highest confidence
- Return type:
numpy.ndarray
- set_adjacent_layers(previous_layer, next_layer)#
Set adjacent layers which are needed for the model to iterate through the layers.
- Parameters:
previous_layer (_Layer) – Layer that is previous to this layer.
next_layer (_Layer) – Layer that is subsequent to this layer.