Author : Ziweritin, stanley
Keyword : Neural network, data mining, support vector machine, pixilated digit, recognition
Subject : Computer science
Article Type : Others
DOI : 10.14445/23488387/ijcse-v7i10p102
Article File : Full Text PDF
Abstract : Recognition is one of the major areas that have attracted the attention of different researchers, which can be applied in every sphere of life as technology advances. There are several problem domains in adopting data mining classification models with the rise in an exponential growth of structured and unstructured data. High metrics of success rate has not been recorded despite the existence and usefulness of data mining classification models in practice. Especially in the areas of testing and training of classifiers to recognize digits on pixilated images like the existing methods which are not efficient and encouraging in terms of speed and accuracy. Because of the segmented colour grid arrangements or formation of some digits. Therefore; we adopted the proposed model to overcome the challenges facing the methods on pixilated digit recognition systems. The aim is to build an efficient pixilated digit recognition system using neural network and support vector machine data mining classification models which can recognize digits within the range of 0-to-9 inclusively from pixilated or raster images. The system was successfully trained and tested in comparison to ascertain 94% and 99% accuracy level for support vector machine and neural network models respectively using Python programming language
Article by : stanley ziweritin
Article add date : 2020-11-12
How to cite : Ziweritin, stanley. (2020-November-12). Building data mining classification model for pixilated digit recognition system. retrieved from https://openacessjournal.com/abstract/361