Different activation function based on input-CodePudding

I am trying to build a Keras neural network where the activation function at the output layer (conditionally) depends on the inputs. The activation function is quite complicated, so as a simpler example, consider the following:

def myactivation(y,x1,x2):
    if x1 > 1.2:
        return(tf.constant(0.0))
    else:
        return(tf.math.minimum(y,x2))

The network consists of an input layer (i.e. x1, x2), a hidden layer with output y and an output layer. The activation above corresponds to the activation of the output layer.

How would I implement something like this? Grateful for any help and guidance you might have!

CodePudding user response：

You can write a custom layer like shown in the keras docs. For your application, the key is to use tf.where based on your condition x1 > 1.2 to switch between the two activation types. Regarding "organizing" the input to the layer, I've decided to stitch them together using tf.stack, such that there is a feature axis in the resulting tensor inp (you'll see in the example).

import tensorflow as tf
import keras

# For reference: https://keras.io/guides/making_new_layers_and_models_via_subclassing/
class AwesomeLayer(keras.layers.Layer):
    def __init__(self, threshold): # Here you could define the layer shape if needed
        super(AwesomeLayer, self).__init__()
        # Here you could define layer (trainable) parameters
        
        self.threshold = tf.constant(threshold)

    def call(self, inputs):
        return tf.where(
            inputs[..., 0] > self.threshold, 
            tf.constant(0.0), 
            tf.math.minimum(inputs[..., 2], inputs[..., 1])
        )

Let's test if it does what you want:

awesome_layer = AwesomeLayer(1.2)

# Input Example 2

x_1 = tf.constant(2.) # Case x_1 > 1.2
x_2 = tf.constant(-5.)
y = tf.constant(1.)

inp = tf.stack([x_1, x_2, y], axis=-1)
print(inp.shape) # (3,)

outp = awesome_layer(inp)
print(outp.numpy()) # 0.0

# Input Example 2

x_1 = tf.constant(-10.) # Case x_1 < 1.2
x_2 = tf.constant(-5.)
y = tf.constant(1.)

inp = tf.stack([x_1, x_2, y], axis=-1)
print(inp.shape) # (3,)

outp = awesome_layer(inp)

print(outp.numpy()) # -5.0

# Input Example 3

x_1 = tf.constant(-10.) # Case x_1 < 1.2
x_2 = tf.constant(25.)
y = tf.constant(1.)

inp = tf.stack([x_1, x_2, y], axis=-1)
print(inp.shape) # (3,)

outp = awesome_layer(inp)

print(outp.numpy()) # 1.0