Perbandingan Akurasi Algoritma K-Nearest Neighbors (KNN) dan Naive Pattern Search (NPS) dalam Website Handwritten Recognition untuk Latihan Menulis Bentuk
Keywords:
handwritten, recognition, KNN, NPSAbstract
Handwritten recognition is becoming increasingly relevant in the current era of technological development. In this article, we discuss the comparison of the accuracy of the KNN and NPS algorithms used in developing handwritten recognition websites specifically for practicing writing shapes for children aged 4–7 years. comparison between the two algorithms used in the recognition of handwritten numbers, namely the K-Nearest Neighbors (KNN) and Naive Pattern Search (NPS) algorithms. KNN is a popular classification algorithm that uses the majority of nearest neighbor categories to classify test data. Meanwhile, NPS is a simple pattern matching algorithm that looks for similarities to reference examples in sample data. The results of the comparison of the two algorithms show that both of them provide good results in handwriting recognition. The accuracy achieved by the KNN algorithm reaches 70%, while the NPS reaches 68%. Although the difference in accuracy is not significant, it can be influenced by variations in the handwriting style of each student who tries exercises on the website.