Perbandingan Akurasi Kinerja Pengklasifikasian Citra Bunga Kertas (Bougainvillea) Menggunakan Neural Network dan Random Forest
Total View This Week0
Total View This Week0
Institusion
Universitas Katolik Widya Mandira Kupang
Author
ELISABETH, Noviana
Subject
NK Decorative arts Applied arts Decoration and ornament
Datestamp
2024-05-22 01:56:36
Abstract :
Bougainvillea is an ornamental plant that is easy to find in home gardens and various places.In this research, bougainvillea images are an interesting object to classify by comparing two methods, namely Neural Network and Random Forest. The purpose of comparing the two methods is to find out the comparative performance of the two methods. The dataset with a total of 5200 images is divided into two, namely training and testing sample data with a ratio of 90%:10%. The two machine learning algorithms used in bougainvillea image classifiers are Neural Network and Random Forest. There are also 2 testing methods, 20-fold cross-validation and random sampling. Calculation results from the 20-fold cross-validation Neural Network method with Precision (99,8%), Recall (99,8%), F1-Score (99,8%), and Accuracy (99,8%) values while Random Forest with Precision (95,3%), Recall (95,3%), F1- Score (95,3%) and Accuracy (95,3%) values.Random sampling testing Neural Network method with Precision (99,8%), Recall (99,8%), F1-Score (99,8%) and Accuracy (99,8%) values while Random Forest with Precision values (94,8%),Recall (94,8%), F1-Score (94,8%) and Accuracy (94,8%)were successfully compared with the highest accuracy, namely the 20-fold cross validation Neural Network method with an accuracy value of 99,8%.
Bougainvillea is an ornamental plant that is easy to find in home gardens and various places.In this research, bougainvillea images are an interesting object to classify by comparing two methods, namely Neural Network and Random Forest. The purpose of comparing the two methods is to find out the comparative performance of the two methods. The dataset with a total of 5200 images is divided into two, namely training and testing sample data with a ratio of 90%:10%. The two machine learning algorithms used in bougainvillea image classifiers are Neural Network and Random Forest. There are also 2 testing methods, 20-fold cross-validation and random sampling. Calculation results from the 20-fold cross-validation Neural Network method with Precision (99,8%), Recall (99,8%), F1-Score (99,8%), and Accuracy (99,8%) values while Random Forest with Precision (95,3%), Recall (95,3%), F1- Score (95,3%) and Accuracy (95,3%) values.Random sampling testing Neural Network method with Precision (99,8%), Recall (99,8%), F1-Score (99,8%) and Accuracy (99,8%) values while Random Forest with Precision values (94,8%),Recall (94,8%), F1-Score (94,8%) and Accuracy (94,8%)were successfully compared with the highest accuracy, namely the 20-fold cross validation Neural Network method with an accuracy value of 99,8%.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
