Comparative Analysis: Handwritten Character Recognition Via K-Nearest Neighbour And Naive Bayes Algorithms

Abstract

The Study of Handwritten Character Prediction: Comparative Analysis Between K-Nearest Neighbour and Naive Bayes Algorithms This research aims to predict handwritten characters by employing the K-Nearest Neighbour (KNN) algorithm and comparing its feature extraction precision with the Naive Bayes (NB) algorithm. Both algorithms were evaluated using a sample size of n = 25 and iterated 15 times to enhance accuracy and reduce errors, with a statistical power of 80% (G power) and an alpha value of 0.05.The experimental results demonstrated a significant improvement in the accuracy of handwritten character recognition, achieving a significance value of 0.165 through the comparison of NB and KNN algorithms. Comparative analysis revealed that the NB algorithm (97.45) outperformed the KNN algorithm (97.3). These findings highlight the superiority of the NB algorithm in predicting handwritten characters over the KNN algorithm.