Klasifikasi Jenis Kucing Menggunakan Algoritma K-Nearest Neighbor

Abstract

Cats in Latin is Felis silvestris catus is a kind of carnivorous animal. Cats are the most popular pets in the world that have many enthusiasts and fans. Cats that have bloodlines are officially registered as purebred cats or pure breeds. The number of cat breeds like this is only slightly, only 1% of the world's cat population, which is usually only bred in official animal husbandry. This study uses the Principal Component Analysis (PCA) and K-nearest Neighbor (KNN) algorithms with the aim of classifying cat images through the analysis stage on original images, binary images and grayscale images.The output of the feature extraction will be the input for the Principal Component Analysis (PCA) and K-nearest Neighbor (KNN) algorithms for cat species classification applications. The feature extraction that will be used in this research are RGB and HSV. The data that will be used in this study are 34 image data, consisting of 24 training data images and 10 test image data.So with this research, it is hoped that it can help people to more easily find out the classification of pets, namely cats. The output accuracy in the classification application uses the Multi Support Vector Machine (SVM) Algorithm with first- order feature extraction from the Principal Component Analysis (PCA) and K-nearest Neighbor (KNN) algorithms, which reaches an accuracy rate of 80%.

Keywords

Pure Breed ; K-NN ; PCA ; RGB ; HSV