Sistem Prediksi Kualitas Santan Kelapa Menggunakan Nearest Mean Classifier (Nmc)

Abstract

Coconut milk is one of the main ingredients that is always used as an ingredient for all types of food. Quality is crucial in choosing coconut milk. However, identifying the quality of coconut milk in plain view is not efficient. It occurs because it is difficult to distinguish which coconut milk is pure and which one is mixed with water. The purpose of this study was to build a quality detection system for coconut milk-based on the color of the coconut milk. The classification algorithm used is the nearest mean classifier (NMC). This method calculates the distance of the image input vector to each class mean of the training image. The closest distance is the basis for determining the results of the classification. The evaluation uses the holdout validation method using a total of 135 images with a ratio of 2/3 for sample data and 1/3 for test data. The evaluation was carried out using 3 types of smartphone cameras, namely 1 Xiaomi Mi 8 Lite, 2 Oppo F7 cameras, and 3 Samsung Galaxy J3 Pro cameras. In testing, the first camera has an accuracy rate of 86.66% compared to camera 2 with an accuracy of 60% and camera 3 with an accuracy of 46%.

Klasifikasi Kopra Putih Menggunakan Algoritma k-Nearest Neighbor

“…Hasil yang diperoleh ini tentunya belum memuaskan. Sistem lainnya adalah sistem prediksi kualitas santan [4]. Tujuan penelitian membangun sistem prediksi kualitas santan berdasarkan warna menggunakan nearest mean classifier (NMC).…”

Metode Fuzzy Mamdani Pada Sistem Pendukung Keputusan Penentuan Jumlah Petugas BPBD Takengon

“…Beberapa penelitian menggunakan metode Fuzzy mamdani sudah pernah dilakukan. Meskipun tujuannya sama untuk memprediksi namun objek yang diteliti berbeda-beda [7]. Metode Fuzzy mamdani juga dilakukan untuk prediksi suatu kegiatan yang memperkirakan apa yang terjadi pada masa akan datang [8].…”