In this work, we carry out a first exploration of the possibility of increasing the performance of Deep Neural Networks (DNNs) by applying diversity techniques to them. Since DNNs are usually very strong, weakening them can be important for this purpose. This paper includes experimental evidence of the effectiveness of binarizing multi-class problems to make beneficial the application of bagging to Denoising Auto-Encoding-Based DNNs for solving the classical MNIST problem. Many research opportunities appear following the diversification idea: We mention some of the most relevant lines at the end of this contribution.
Paper
Full text
Does diversity improve deep learning?
Semantic Scholar · Computer Science · 2015
Abstract
In this work, we carry out a first exploration of the possibility of increasing the performance of Deep Neural Networks (DNNs) by applying diversity techniques to them. Since DNNs are usually very strong, weakening them can be important for this purpose. This paper includes experimental evidence of the effectiveness of binarizing multi-class problems to make beneficial the application of bagging to Denoising Auto-Encoding-Based DNNs for solving the classical MNIST problem. Many research opportunities appear following the diversification idea: We mention some of the most relevant lines at the end of this contribution.