Exponential Linear Unit (ELU) is a popular activation function that speeds up learning and produces more accurate results. This article is an introduction to ELU and its position when compared to other popular activation functions. It also includes an interactive example and usage with PyTorch and Tensorflow.
Djork-Arné Clevert, Thomas Unterthiner, Sepp Hochreiter introduced ELU in Nov 2015. It outperformed ReLU-based CIFAR-100 networks at the time. To this day, ELUs are still popular among Machine Learning engineers and are well studied by now.
ELU is an activation function based on ReLU that has an extra alpha constant (α) that defines function smoothness when inputs are negative. Play with an interactive example below to understand how α influences the curve for the negative part of the function.
The ELU output for positive input is the input (identity). If the input is negative, the output curve is slightly smoothed towards the alpha constant (α). The higher the alpha constant, the more negative the output for negative inputs gets.
ELU and ReLU are the most popular activation functions used. Here are the advantages and disadvantages of using it when compared to other popular activation functions.
- Tend to converge faster than ReLU (because mean ELU activations are closer to zero)
- Better generalization performance than ReLU
- Fully continuous
- Fully differentiable
- Does not have a vanishing gradients problem
- Does not have an exploding gradients problem
- Does not have a dead relu problem
- Slower to compute (because of non-linearity for negative input values)
ELU is slower to compute, but ELU compensates this by faster convergence during training. During test time ELU is slower to compute than ReLU though.
To use ELU in PyTorch, use
torch.nn.ELU function. Here's an example model that uses ELU:
import torch from torch import nn class Model(nn.Module): def __init__(self, dataset): super(Model, self).__init__() self.layer1 = nn.Sequential( nn.Conv3d(in_channels=4, out_channels=2, kernel_size=2), nn.ELU(alpha=2.0) ) def forward(self, x): return self.layer1(x)
1 by default. To learn more, read PyTorch ELU documentation.
The easiest way to use ELU in TensorFlow is to use Keras layers. Example TensorFlow model below that uses ELU:
import tensorflow as tf class Model(tf.keras.Model): def __init__(self): super(Model, self).__init__() self.layer1 = tf.keras.models.Sequential([ tf.keras.layers.Conv2D(4, kernel_size=(2, 2), input_shape=(6, 6, 1)), tf.keras.layers.ELU(alpha=2.0) ]) def call(self, x): return self.layer1(x)
1 by default. More details in TensorFlow ELU documentation.
Now you know how ELU works and where it stands compared to other popular activation functions.
Here's what you can do next: