Author

Digya Acharya

Date of Award

2022

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Computer Science

Committee Chair

Chaity Banerjee

Committee Member

Letha Etzkorn

Committee Member

Vineetha Menon

Subject(s)

Deep learning (Machine learning), Pattern recognition systems, Image processing

Abstract

In this thesis we address the problem of consensus based multi-label classification. In the general multi-label classification problem, an input is assigned to multiple classes without any constraints. However, there are applications where the multi-label classification needs to be solved under the constraint of a consensus among the assigned labels. We address this problem here using a deep learning approach. We conduct experiments with the MNIST dataset and establish the possibility of using such approaches for traditional multi-class problems. Furthermore, we demonstrate the use of our consensus based classifier in a generative adversarial framework for training a generator for handwritten digits. Our approach results in a lower loss and better quality generated images.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.