Optimization of Hyperparameter Tuning in the Support Vector Machine Method for Classifying Coffee Fruit Maturity Levels
-
Juprianus Rusman(1*)
Universitas Kristen Indonesia Toraja
-
Berlian Zetikarya Haryati(2)
Universitas Kristen Indonesia Toraja
-
Aryo Michael(3)
Universitas Kristen Indonesia Toraja
(*) Corresponding Author
Abstract
This study aims to develop an optimal classification model to identify the maturity level of coffee fruit based on color features of coffee fruit images. The color features red, green and blue in RGB color space and hue, saturation and value in HSV color space are extracted from the coffee fruit image and used as input for the support vector machine (SVM) classification model. In order to optimize the performance of SVM, hyperparameter tuning is used with the grid search method to determine the best parameters in the classification model built. By using 180 training images in determining the optimum parameters, the hyperparameter tuning results of the grid search method are obtained at cross validation (cv) = 6, cost (C) = 1000, gamma (γ) = 0.001, and kernel = linear. Then the optimum parameters are used as an SVM model to classify 45 test images into three different classes, namely ripe, half ripe and raw. Based on the evaluation with confusion matrix, it can be concluded that the built model has good performance with 98% accuracy. This indicates that the model is able to distinguish the three classes with a low error rate. With this ability, the model built has great potential in the agricultural industry to support the use of technology in agriculture, especially coffee farming such as post-harvest fruit processing. This model can be developed on a maturity level sorting machine.
Downloads
References
E. Prakasa, ‘Computer Vision dan Artificial Intelligence di Era Revolusi Industri 4.0 - General Lecture in Universitas Riau 28 October 2021’, Oct. 28, 2021 [Online]. Available: 10.13140/RG.2.2.26229.09448.
H. Fitriyah and R. C. Wihandika, Dasar-Dasar Pengolahan Citra Digital. Universitas Brawijaya Press, 2021.
M. Z. Andrekha and Y. Huda, ‘Deteksi Warna Manggis Menggunakan Pengolahan Citra dengan Opencv Python’, Voteteknika Vocat. Tek. Elektron. Dan Inform., vol. 9, no. 4, pp. 27–33, Dec. 2021 [Online]. Available: 10.24036/voteteknika.v9i4.114251.
Seema et al., ‘Automatic Fruit Grading and Classification System Using Computer Vision: A Review’, in 2015 Second International Conference on Advances in Computing and Communication Engineering, 2015, pp. 598–603 [Online]. Available: 10.1109/ICACCE.2015.15.
N. Astrianda, ‘Klasifikasi Kematangan Buah Tomat Dengan Variasi Model Warna Menggunakan Support Vector Machine’, VOCATECH Vocat. Educ. Technol. J., vol. 1, pp. 45–52, Apr. 2020 [Online]. Available: 10.38038/vocatech.v1i2.27.
I. Indrabayu et al., ‘Strawberry Ripeness Classification System Based On Skin Tone Color using Multi-Class Support Vector Machine’, in 2019 International Conference on Information and Communications Technology (ICOIACT), 2019, pp. 191–195 [Online]. Available: 10.1109/ICOIACT46704.2019.8938457.
K. Kr et al., ‘An Improved Feature Selection and Classification of Gene Expression Profile using SVM’, in 2019 2nd International Conference on Intelligent Computing, Instrumentation and Control Technologies (ICICICT), 2019, vol. 1, pp. 1033–1037 [Online]. Available: 10.1109/ICICICT46008.2019.8993358.
M. Fajri and A. Primajaya, ‘Komparasi Teknik Hyperparameter Optimization pada SVM untuk Permasalahan Klasifikasi dengan Menggunakan Grid Search dan Random Search’, J. Appl. Inform. Comput., vol. 7, no. 1, pp. 10–15, Jul. 2023 [Online]. Available: 10.30871/jaic.v7i1.5004.
P. C. Upadhyay et al., ‘Classifying Cover Crop Residue from RGB Images: A Simple SVM versus a SVM Ensemble’, in 2021 IEEE Symposium Series on Computational Intelligence (SSCI), 2021, pp. 1–7 [Online]. Available: 10.1109/SSCI50451.2021.9660147.
I. S. Areni et al., ‘Klasifikasi Kematangan Stroberi Berbasis Segmentasi Warna dengan Metode HSV’, J. Penelit. Enj., vol. 23, no. 2, pp. 113–116, Nov. 2019 [Online]. Available: 10.25042/jpe.112019.03.
M. M. Chandra and Yoannita, ‘Klasifikasi Jenis Bunga Menggunakan Metode Svm Berdasarkan Citra Dengan Fitur Hsv’, J. Indones. Sos. Teknol., vol. 4, no. 02, pp. 255–264, Feb. 2023 [Online]. Available: 10.59141/jist.v4i02.585.
Y. Dong et al., ‘A Machine Learning Model for Product Fraud Detection Based On SVM’, in 2021 2nd International Conference on Education, Knowledge and Information Management (ICEKIM), 2021, pp. 385–388 [Online]. Available: 10.1109/ICEKIM52309.2021.00091.
Y. E. Yana and N. Nafi’iyah, ‘Klasifikasi Jenis Pisang Berdasarkan Fitur Warna, Tekstur, Bentuk Citra Menggunakan SVM dan KNN’, Res. J. Comput. Inf. Syst. Technol. Manag., vol. 4, no. 1, pp. 28–36, Apr. 2021 [Online]. Available: 10.25273/research.v4i1.6687.
Z. M. E. Darmawan and A. F. Dianta, ‘Implementasi Optimasi Hyperparameter GridSearchCV Pada Sistem Prediksi Serangan Jantung Menggunakan SVM’, Teknol. J. Ilm. Sist. Inf., vol. 13, no. 1, pp. 8–15, Jan. 2023 [Online]. Available: 10.26594/teknologi.v13i1.3098.
A. G. Sooai et al., ‘Klasifikasi Citra Daun Anggur Menggunakan SVM Kernel Linear’, JOINTECS J. Inf. Technol. Comput. Sci., vol. 8, no. 1, pp. 19–26, Mar. 2023 [Online]. Available: 10.31328/jointecs.v8i1.4496.
Copyright (c) 2023 Juprianus Rusman, Berlian Zetikarya Haryati, Aryo Michael
This work is licensed under a Creative Commons Attribution 4.0 International License.
The author submitting the manuscript must understand and agree that if accepted for publication, authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution (CC-BY) 4.0 License that allows others to share the work with an acknowledgment of the work’s authorship and initial publication in this journal.
