Handwritten Identification Using K Nearest Neighbor and Compared with Feature Extraction Precision Naive Bayes Algorithm

The primary goal of the research is to examine handwriting identification using the k closest neighbor (KNN) algorithm and evaluate how accurate it is by contrasting its feature precision with that of the Naive Bayes (NB) approach. The Naive Bayes (NB) algorithm (N=10) and the k nearest neighbor (N=10) with a g-power value of 80% were used in the proposed work to identify handwriting. Datasets with thresholds of 0.05 %, confidence intervals of 95 % mean and standard deviation, and recent study findings were gathered from various web sources. was run through twenty times to gather data. The handwritten character had the highest level of precision. When compared to the k nearest neighbor algorithm, the identification using Naive Bayes is (97.45%), with minimum mean error of 97.3%, and the significance value is 0.03 (p < 0.05 Independent sample T-test). For handwriting identification, the Naive Bayes algorithm performed more accurately and classified the digits more well.