Unrestricted Online Kannada Handwriting Recognition using Data Mining Techniques

Ramya, S and Shama, Kumara and Niranjan, U C (2017) Unrestricted Online Kannada Handwriting Recognition using Data Mining Techniques. In: ICRTESME, 08/08/2017, MIT Manipal.

[img] PDF
833.pdf - Published Version
Restricted to Registered users only

Download (778kB) | Request a copy


This article presents an empirical assessment of the effectiveness of various Data mining techniques on unrestricted online Kannada handwriting recognition. Twelve different classifiers are implemented, and performance evaluation is carried out based on the recognition accuracy, root-mean-square error, precision, recall and F-measure. Two separate training and testing methodologies namely, single partition and ten-fold cross-validation methods are implemented and tested. In this work, 5000 samples of isolated online Kannada handwritten character data are collected from 62 different writers. A novel framework for online handwritten Kannada character recognition using Data mining classification methods is presented which provides the innovative benchmark for future research. Recognition accuracy of 96.6% is achieved with ten-fold cross validation method and K-Nearest Neighbour classifier

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: Online Handwriting Recognition; Data mining; Preprocessing; Feature Extraction, Pattern Recognition.
Subjects: Engineering > MIT Manipal > Electronics and Communication
Depositing User: MIT Library
Date Deposited: 08 Sep 2017 10:15
Last Modified: 08 Sep 2017 10:15
URI: http://eprints.manipal.edu/id/eprint/149653

Actions (login required)

View Item View Item