What is it about?
Proposing a new methodology to prune the neural network based on input performance. The approach utilizes the nature of the input data and its importance, the more is the importance of the input data, the more likely is to keep its input link to hidden layer, and the lower is the importance of the input data, the less likely is to keep its input link to hidden layer.
Featured Image
Photo by Hitesh Choudhary on Unsplash
Why is it important?
neural networks nowadays are used commonly in machine learning. Due to complexity of neural networks, the learning process takes a lot of time, one way to expedite the learning process is to minimize the size of the neural network. This process is done by what so called pruning.
Perspectives
I am sure that this work will help and promote other researchers to continue in this research field. I think it will open avenues of research because the used approach is unique of its mechanism by using the input importance.
Nabil Hewahi
University of Bahrain
Read the Original
This page is a summary of: Neural network pruning based on input importance, Journal of Intelligent & Fuzzy Systems, September 2019, IOS Press,
DOI: 10.3233/jifs-182544.
You can read the full text:
Contributors
The following have contributed to this page
