MODEL IDENTIFIKASI MANGGA MATANG ALAMI MENGGUNAKAN LEARNING VECTOR QUANTIZATION (LVQ)
Total View This Week0
Total View This Week0
Institusion
Universitas Pembangunan Nasional Veteran Jakarta
Author
Luthfi Khalid, .
Subject
T Technology (General)
Datestamp
2022-01-13 02:23:18
Abstract :
Mature mangoes are classified into 2 parts, naturally and ripe using calcium carbide. The use of calcium carbide compounds can accelerate the ripening of mangoes. This study aims to classify natural ripe mangoes with ripe mangoes using calcium carbide. The data used in this study are primary data images taken using a smartphone. To distinguish the characteristics of natural or ripe mangoes using calcium carbide we used RGB images in this study. The Learning Vector Quantization (LVQ) algorithm is used to classify natural or ripe mature mangoes using calcium carbide. Feature extraction uses mean, variance, standard deviation. The best accuracy obtained in the training process uses 24 image data with a hidden size of 25 learning rate 0.1 and an error goal of 0.01, an accuracy of 95.8333% is obtained. In the testing process using 16 image data, 87.5% accuracy is obtained, with a learning rate of 0.1, hidden size 25 and error goal 0.01.
Mature mangoes are classified into 2 parts, naturally and ripe using calcium carbide. The use of calcium carbide compounds can accelerate the ripening of mangoes. This study aims to classify natural ripe mangoes with ripe mangoes using calcium carbide. The data used in this study are primary data images taken using a smartphone. To distinguish the characteristics of natural or ripe mangoes using calcium carbide we used RGB images in this study. The Learning Vector Quantization (LVQ) algorithm is used to classify natural or ripe mature mangoes using calcium carbide. Feature extraction uses mean, variance, standard deviation. The best accuracy obtained in the training process uses 24 image data with a hidden size of 25 learning rate 0.1 and an error goal of 0.01, an accuracy of 95.8333% is obtained. In the testing process using 16 image data, 87.5% accuracy is obtained, with a learning rate of 0.1, hidden size 25 and error goal 0.01.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
