Digit Classification Using Signatures

Conference Year

January 2022

Abstract

The goal of classification using signatures is to improve the efficiency of neural networks and other classifiers by using less data to achieve the same or better level of accuracy. By calculating the signatures of hand drawn digits, a larger photo containing thousands of pixels can be represented by far fewer values that still allow the photo to be classified. Digits are used to showcase the benefits of classification using signatures by easily drawing a digit of your own to test the accuracy of this classification system.

Primary Faculty Mentor Name

Luis Duffaut Espinosa

Status

Undergraduate

Student College

College of Engineering and Mathematical Sciences

Program/Major

Electrical Engineering

Primary Research Category

Engineering & Physical Sciences

Abstract only.

Share

COinS
 

Digit Classification Using Signatures

The goal of classification using signatures is to improve the efficiency of neural networks and other classifiers by using less data to achieve the same or better level of accuracy. By calculating the signatures of hand drawn digits, a larger photo containing thousands of pixels can be represented by far fewer values that still allow the photo to be classified. Digits are used to showcase the benefits of classification using signatures by easily drawing a digit of your own to test the accuracy of this classification system.