Novel Analog Implementation of a Hyperbolic Tangent Neuron in Artificial Neural Networks

Fatemeh Mohammadi Shakiba, Mengchu Zhou

Research output: Contribution to journalArticlepeer-review

Abstract

Recently, enormous datasets have made power dissipation and area usage lie at the heart of designs for an Artificial Neural Network (ANN). Considering the significant role of activation functions in the neurons and the growth of hardware-based neural networks like Memristive Neural Networks, this work proposes a novel design for a hyperbolic tangent activation function (Tanh) to be used in memristive-based neuromorphic architectures. The purpose of the implementation of a CMOS-based design for Tanh is to decrease power dissipation and area usage. This design also increases the overall speed of computations in ANNs, while keeping the accuracy in an acceptable range. The proposed design is the first analog one for the hyperbolic tangent and its performance is analyzed by using well-known Modified National Institute of Standards and Technology (MNIST) dataset and Fashion-MNIST one.

Original languageEnglish (US)
JournalIEEE Transactions on Industrial Electronics
DOIs
StateAccepted/In press - 2020

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Keywords

  • Activation function
  • Artificial Neural network (ANN)
  • Biological neural networks
  • Computer architecture
  • Hardware
  • Hardware implementation
  • Hyperbolic tangent
  • Memristive Neural Network (MNN)
  • Memristors
  • Multi-layer neural network
  • Neuromorphic architecture
  • Neurons
  • Power demand

Fingerprint Dive into the research topics of 'Novel Analog Implementation of a Hyperbolic Tangent Neuron in Artificial Neural Networks'. Together they form a unique fingerprint.

Cite this