Handwritten digit classification using novel support vector machine algorithm comparing with KNN for improving accuracy

To forecast the categorization of digital handwriting images, this study employs two classification algorithms: K-Nearest Neighbors (KNN) and Nova Support Vector Machine (SVM). Using a Digit dataset consisting of 935 items, the classification method is evaluated. Our suggested method makes use of support vector machines (SVMs) and kernel neural networks (KNNs) to predict numerical values typed by hand. Using G power, we split the staff into two 32-person teams. The clinical trials sample size was determined using the following criteria: 90 percent confidence, 0.5 beta, alpha of 0.05, pre-test G power of 80%, enrollment ratio of 1, and beta of 0.5. Compared to KNN, the Novel Support Vector Machine classifier has a success rate of 94.01% when predicting the Handwritten Digit classification on the dataset. There is a notable difference between the two groups, as shown by a p-value of 0.002 (p