Performance Analysis of Indo-Aryan text classification using ALEX net and VGG-16 Deep Learning Architecture

Abstract

Text classification is utilized to put together character in a predefined set of classes. It is exceptionally helpful in Web content administration, web crawlers, email sifting, and so forth. Text classification is a troublesome errand because of high-dimensional element vector containing glitzy and unimportant features. Different feature cut techniques have been already proposed for dispensing with superfluous features just as for lessening the element of vector. Pertinent and diminished element vector is utilized by AI model for better classification results. This paper compares the performance of two different text classification approaches for Indo Aryan language by utilizing AI methods. From the comparison it will be very clear that VGG-16 net architecture performs better for classification of Indo Aryan character images.

Keywords-Alex net, VGG-16, Indo Aryan, text classifcation, deep learing, and Saurashtra text