Claim Missing Document
Check
Articles

Found 2 Documents
Search

Penerapan Metode K-Nearest Neighbor dan Support Vector Machine untuk Klasifikasi Kematangan Buah Mengkudu Hehanussa, Siti Gayatri; Hartono Wijaya, Sony; Haryanto, Toto
Jurnal Ilmu Komputer dan Agri-Informatika Vol 12 No 1 (2025)
Publisher : Departemen Ilmu Komputer, Institut Pertanian Bogor

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29244/jika.12.1.25-37

Abstract

Noni fruit (Morinda citrifolia) is one of Indonesia’s export commodities. It is available year-round and is well known for its numerous health benefits. Native to Southeast Asia, including Indonesia, noni fruit is widely used in traditional medicine. Typically, the ripeness of noni fruit is determined manually based on visual inspection, which can lead to subjective judgments and inconsistent results. Therefore, this study aims to develop a machine-learning model to classify the ripeness levels of noni fruit. The classification methods employed are K-Nearest Neighbor (KNN) and Support Vector Machine (SVM), utilizing Hue-Saturation-Intensity (HSI) color features and Local Binary Pattern (LBP) texture features. Experimental results show that the KNN algorithm outperforms the SVM algorithm in terms of classification accuracy. The highest accuracy achieved using KNN was 88.62% at k = 11, whereas the best accuracy obtained with SVM using a polynomial kernel was 87.80%, with parameters set to C = 0.1, Gamma = 1, Degree = 5, and coef0 = 1.0. These results were achieved using an 80:20 split ratio for training and testing data.
Pengembangan Model Multilayer Classifier Menggunakan Metode Ensemble Learning untuk Grading Brokoli Imaduddin, Zaki; Purwanto, Yohanes Aris; Hartono Wijaya, Sony; Nidya Neyman, Shelvie
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 12 No 4: Agustus 2025
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.124

Abstract

Permintaan brokoli di Indonesia terus meningkat 15% sampai dengan 20% per tahun. Sayangnya supply masih terbatas dan kualitas masih kurang. Untuk menentukan kualitas brokoli diperlukan adanya proses grading yaitu proses pemeringkat brokoli menjadi grade A, B dan C berdasarkan tiga parameter utama yaitu warna, ukuran, dan bentuk. Sayangnya, tidak semua petani memahami mengenai proses grading tersebut. Hal ini menyebabkan kerugian pada petani dan pengusaha brokoli. Penelitian ini bertujuan untuk mengembangkan algoritma grading menggunakan Convolusional neural network (CNN) berdasarkan 2 buah citra yaitu citra kepala brokoli dari tampak atas dan tampak samping. Dataset pada penelitian ini sebesar 600 data. Teknik yang digunakan ialah modifikasi beberapa model deep learning yaitu ResNet50, EfficientNetB2, VGG16 pada bagian layer klasifikasinya, lalu dibandingkan dengan hasil akurasi dari masing-masing outputnya. Penelitian ini juga menggunakan metode ensemble learning dimana menggunakan kombinasi dari 3 fitur berbeda. Fitur warna, ukuran dan bentuk digabungkan pada proses training dan testing untuk melakukan klasifikasi grade brokoli. Pada fase testing digunakan teknik voting untuk pengambilan keputusan grading. Akurasi terbaik ada pada model ResNet50 dengan hasil klasifikasi brokoli sebesar 90% yang didapatkan melalui penggunaan 5 dense layer pada layer klasifikasi, sehingga mampu melebihi hasil akurasi dari beberapa model deep learning lainnya. Algoritma dari penelitian ini menawarkan solusi grading yang lebih objektif dan konsisten dibandingkan sistem manual, sehingga petani dan pengusaha brokoli dalam meningkatkan efisiensi, mengurangi kerugian, dan memastikan kualitas produk yang lebih baik bagi konsumen.   Abstract The demand for broccoli in Indonesia has been increasing by 15% to 20% annually. However, supply remains limited, and quality control is inadequate. To assess broccoli quality, a grading process is required, classifying broccoli into Grades A, B, and C based on three primary parameters: color, size, and shape. Unfortunately, not all farmers possess sufficient knowledge of this grading process, leading to financial losses for both farmers and broccoli businesses. This study aims to develop a grading algorithm using a Convolutional Neural Network (CNN) based on two images, namely a top-view and a side-view image of a broccoli head. The dataset comprises 600 samples. The methodology involves modifying the classification layers of several deep learning models, namely ResNet50, EfficientNetB2, and VGG16, and comparing their classification accuracy. Additionally, an ensemble learning approach is employed, integrating three distinct features—color, size, and shape—into the training and testing phases for broccoli grading. The voting technique is utilized in the testing phase to enhance decision-making in the grading process. Experimental results indicate that the ResNet50 model achieves the highest classification accuracy at 90%, attributed to the incorporation of five dense layers in the classification stage. This performance surpasses that of other deep learning models. The proposed algorithm provides a more objective and consistent grading system compared to manual methods, enabling farmers and broccoli enterprises to enhance efficiency, reduce financial losses, and ensure higher product quality for consumers.