Shafiyyatul Amaliyyah School Student Face Absence Using Principal Component Analysis and K – Nearest Neighbor

Abstract

Pattern recognition is one of the sciences used to classify things based on quantitative measurements of the main features or properties of an object. Pattern recognition has been widely used in various fields of research. One of the pattern recognition that is often discussed is facial recognition. The face is one of the human biometrics that is often used as the main information of a person. Face recognition is a field of research with many applications in applications such as attendance, population data collection, security systems, and others. The research utilizes feature extraction of PCA (Principal Component Analysis), and K-NN (K – Nearest Neighbor) with variations of the distance formula by applying facial recognition attendance at the Safiatul Amaliyah School. This research is expected to get accurate results in detecting, recognizing, and comparing a person's face with a small error rate. The distance formula with accuracy level is presented with the equation Cityblock < Euclidian < Minkowski < Chebychev. The effect of applying the variation of the distance formula on the performance of the facial attendance recognition model is not too big, but it is better.

Identification of Tempe Fermentation Maturity Using Principal Component Analysis and K-Nearest Neighbor

“…Near and far neighbors of pixels are calculated by Euclidean distance, while Principal Component Analysis (PCA) is an algorithm that is able to convert a group of data that are initially correlated to data that are not correlated to each other (Principal Component). The number of Principal Components produced is the same as the amount of the original data, but can be reduced to a smaller number and is still able to represent the original data well (Rambe, Tanjung, and Muhathir 2022).…”

“…Gray Level Co-occurrence Matrix (GLCM) is a method used for texture analysis as part of feature extraction. GLCM is a matrix that describes the frequency of the appearance of a pair of two pixels with a certain intensity in a certain distance and direction in the image (Rambe et al 2022).…”

Numerical Analysis of Variations Distance Formulas on K Nearest Neighbors In Classifying Malaria Parasite Blood Cells

“…Struktur Algoritma KNN memiliki langkah menghitung jarak dengan memanfaatkan formula jarak, bawan formula jarak algoritma KNN adalah Euclidian distance. Penelitian (Rambe, Tanjung, & Muhathir, 2022) menerapkan variasi formula jarak Euclidian, City Block, Minkowski dan Chebychev pada KNN dalam aplikasi absensi berbasis wajah pada sekolah Syafiyatul Amaliyah, temuan yang dicapai formula jarak Cityblock menduduki perolehan tingkat akurasi yang paling tinggi kemudian diikuti formula jarak Euclidian formula jarak Minkowski dan formula jarak Chebychev memperoleh hasil akurasi yang minimal dengan perolehan akurasi berturut-turut 78.79%, 77.88%, 76.97%, 67.27%. Formula jarak mempunyai keunikannya tersendiri dalam menentukan jarak terdekatnya, sehingga dalam penelitian ini melanjutkan penelitian terdahulu dengan memanfaatkan lebih banyak formula jarak dalam mengklasifikasi sel darah malaria parasit.…”

Rancang Bangun Aplikasi Kehadiran Siswa Menggunakan Pengenalan Wajah Berbasis Android Dengan Metode Machine Learning

“…Arsitektur dari aplikasi kehadiran siswa menggunakan face recognition [15], [16], [17] dimana dari sisi Back-End: merupakan aplikasi server WSGI [18], [19] yang bertugas sebagai interface dalam menangani penyimpanan dataset wajah user. server menggunakan server environment (.env) dengan library python 3.8, OpenCV Python, Face Recognition, Server Flask, Json dimana user akan melakukan registrasi wajah pada sistem kemudian hasil registrasi akan disimpan pada database wajah sebanyak 51 contoh wajah, jumlah wajah tersebut selanjutnya akan di generate dan dilakukan training data wajah dengan metode KNN [20] (trained_knn_model.clf) untuk menentukan nilai kedekatan dari sampel wajah tersebut dimana nilai kedekatan/akurasi nilai tertentu akan menunjukan akurasi wajah seorang siswa selanjutnya disimpan untuk verifikasi wajah jika ada request dari aplikasi android.…”