Vijayalaxmi Rudraswamimath | VTU Belgaum, Karnataka, INDIA (original) (raw)
Uploads
Papers by Vijayalaxmi Rudraswamimath
In recent years, identification process of handwritten digits has turn out to be an important and... more In recent years, identification process of handwritten digits has turn out to be an important and notable issue. From early age of this system, a measureless methods have intended by research community for improving an accuracy rate of recognition but due to associated issues of offered methods field is still open for research. This paper explores handwritten digits recognition act of five different schemes over MNIST dataset that consist huge handwritten digits. Experimental fallouts denoted that CNN technique is most suitable scheme for recognizing hand written digits, obtained 99.99% accuracy in contrast of ANN, KNN, SVM, and RFC algorithm. Keywords Convolution neural network (CNN) • Artificial neural network (ANN) • Support vector machine (SVM) • K-nearest neighbor (KNN)
In recent years, identification process of handwritten digits has turn out to be an important and... more In recent years, identification process of handwritten digits has turn out to be an important and notable issue. From early age of this system, a measureless methods have intended by research community for improving an accuracy rate of recognition but due to associated issues of offered methods field is still open for research. This paper explores handwritten digits recognition act of five different schemes over MNIST dataset that consist huge handwritten digits. Experimental fallouts denoted that CNN technique is most suitable scheme for recognizing hand written digits, obtained 99.99% accuracy in contrast of ANN, KNN, SVM, and RFC algorithm. Keywords Convolution neural network (CNN) • Artificial neural network (ANN) • Support vector machine (SVM) • K-nearest neighbor (KNN)