hyper-sinh: An Accurate and Reliable Function from Shallow to Deep Learning in TensorFlow and Keras

TL;DR

The paper introduces 'hyper-sinh', a new activation function for neural networks implemented in TensorFlow and Keras, which improves accuracy and reliability in image and text classification tasks across various network depths.

Contribution

It proposes the hyper-sinh activation function, a novel variation of m-arcsinh, validated as effective for both shallow and deep neural networks in supervised learning.

Findings

Hyper-sinh achieves competitive accuracy on benchmark datasets.

It improves reliability in classification tasks.

It performs well in both shallow and deep neural network architectures.

Abstract

This paper presents the 'hyper-sinh', a variation of the m-arcsinh activation function suitable for Deep Learning (DL)-based algorithms for supervised learning, such as Convolutional Neural Networks (CNN). hyper-sinh, developed in the open source Python libraries TensorFlow and Keras, is thus described and validated as an accurate and reliable activation function for both shallow and deep neural networks. Improvements in accuracy and reliability in image and text classification tasks on five (N = 5) benchmark data sets available from Keras are discussed. Experimental results demonstrate the overall competitive classification performance of both shallow and deep neural networks, obtained via this novel function. This function is evaluated with respect to gold standard activation functions, demonstrating its overall competitive accuracy and reliability for both image and text classification.