Feature selection optimization based on genetic algorithm for support vector classification varieties of raisin

Abstract

Grapes are one of the fruit plants that grow that propagate in certain fields. Grapes can be processed into juice, wine, raisins, and so on. Raisins are dried grapes. Raisins have a distinctive taste and aroma. Raisins are a concentrated and nutritious source of carbohydrates, containing antioxidants, potassium, fiber and iron. To increase the accuracy value, the optimize selection genetic algorithm (GA) is used. This research was conducted modeling using the support vector machine (SVM) and SVM algorithms based on optimize selection GA by using the raisin (raisin varieties) dataset obtained from the UCI machine learning repository. The research dataset is divided into training data and testing data. The data sharing will be carried out using the cross validation and split validation operators. Data validation with 10-Fold-validation on the SVM algorithm has the best level of performance among 5 other algorithms such as; Naïve Bayes, K-nearest neighbor (K-NN), decision tree (DT), neural network, and random forest (RF). The SVM algorithm produces accuracy and area under the curve (AUC) values of 87.11% for accuracy and 0.928 for AUC. Optimization in this study using optimize selection GA. SVM based on optimize selection GA produces accuracy and AUC values of 87.67% for accuracy and 0.930 for AUC.

Gesture Recognition untuk Deteksi Bahasa Isyarat BISINDO: Pendekatan Mediapipe dan Random Forest

“…Support Vector Machine (SVM) memiliki keunggulan dalam hal teori yang lengkap, optimasi global, adaptabilitas yang kuat, dan kemampuan generalisasi yang baik karena didasarkan pada Teori Pembelajaran Statistik (Statistical Learning Theory/SLT) [28]. Dalam model SVM, sebuah hyperplane keputusan dibangun untuk membentuk celah pemisahan guna membagi dua contoh kelas dengan margin maksimum [29]. SVM memiliki keunggulan unik dalam menangani pengenalan pola nonlinear, dan dimensi yang tinggi [28].…”

Augmented Rice Plant Disease Detection with Convolutional Neural Networks

“…Several previous studies have conducted rice disease detection using various approaches, such as research [9] using the Random Forest method for medicinal plant disease detection with an accuracy of 98.97%, precision 99.42%, recall 98.89%, and F-measure 99.15%. Research [10] applies genetic feature selection with the SVM method for variety detection to get an accuracy of 87.67% and an AUC of 93%. Research [11] uses the Deep Convolutional Neural Network (DCNN) method to detect rice disease types with precision 0.962, recall 0, 0.9617, specificity 0.9921, and F1-score 0.9616.…”

Improve Coal Blending Optimization in CFPP by Cromosom and Fitness Function Redefinition of the Genetic Algorithm

“…However, GA is computationally expensive and requires careful parameter configuration [6]. GA has demonstrated exemplary performance when implemented in real-world problems, including optimizing CNN architecture with a transfer-learning strategy from parent networks [2], shortest path problem [9],optimizing ANN parameters [10], cryptoanalysis [11], community structure in complex networks [12], multi-objective in packing [13], scheduling [9], combinatorial configuration optimization [5], feature selection Ramdhani 2023 [14], intrution detection suhaimi [15]. There are at least five variants of genetic algorithms, namely real and binary-coded, multiobjective, parallel, chaotic, and hybrid GAs [8].…”