Claim Missing Document
Check
Articles

Found 31 Documents
Search

Optimisasi Model Deep Learning untuk Deteksi Penyakit Daun Tebu dengan Fine-Tuning MobileNetV2 Aryanti, Riska; Agustiani, Sarifah; Wildah, Siti Khotimatul; Arifin, Yosep Tajul; Marlina, Siti; Misriati, Titik
Journal of Informatics Management and Information Technology Vol. 4 No. 4 (2024): October 2024
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/jimat.v4i4.411

Abstract

Sugarcane leaf diseases are a serious threat in sugarcane farming because they can significantly reduce productivity and can cause major losses in yields if not detected early. Therefore, fast and accurate disease management is needed to prevent further losses. This study aims to develop a deep learning model based on MobileNetV2 with fine-tuning techniques to effectively detect sugarcane leaf diseases. Fine-tuning is a method used to adjust the parameters of a pre-trained model on a more specific target dataset. The dataset contains images of sugarcane leaves that have been classified per class based on the type of disease. In this study, fine-tuning was performed on the MobileNetV2 architecture that had been previously trained using the sugarcane leaf dataset. The fine-tuning process was carried out by rearranging the top few layers of MobileNetV2 and adding a special classification layer to predict the class of sugarcane leaf diseases. The model was trained through two stages: initial training to obtain a baseline performance and fine-tuning by opening several layers of MobileNetV2. In the initial evaluation, the model achieved a validation accuracy of 93.12%. After fine-tuning, the accuracy increased to 95.01%, indicating that this technique was able to significantly improve disease detection capabilities. The results of this study provide important contributions in the field of agriculture, especially in supporting the sustainability of sugarcane production through artificial intelligence-based technology. The implementation of the proposed model is expected to help farmers detect diseases more quickly and take timely preventive measures, thereby reducing losses.
Classification of eucalyptus leaves: Combining color histogram feature extraction and decission tree algorithm Agustiani, Sarifah; Hidayat, Rahmat; Arifin, Yoseph Tajul; Haryani; Marlina, Siti
Jurnal Teknik Informatika C.I.T Medicom Vol 16 No 2 (2024): May: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/cit.Vol16.2024.731.pp58-69

Abstract

This research proposes an automatic approach to identify eucalyptus species based on leaf images using color histogram feature extraction and the Decision Tree algorithm. Eucalyptus is known as one of the most productive plants in the world with various uses in the timber, biofuel and pharmaceutical industries. However, its wide environmental adaptability and rapid growth pose challenges in identification and management. The proposed approach focuses on the use of Artificial Intelligence (AI) technology and image analysis to solve the identification problem. The color histogram feature extraction method is used to extract visual information about the color distribution of eucalyptus leaves. The Decision Tree algorithm is used to build a classification model based on the extracted features. Model evaluation is carried out using accuracy, precision, recall and F1-score metrics. The results showed that this approach was effective in identifying eucalyptus species, with a high level of accuracy. In addition, the development of this method offers opportunities for further applications in various fields, including forest mapping, mobile applications, and the timber industry. By combining advances in AI and image analysis, this research has the potential to become an important cornerstone of nature conservation and environmental sustainability efforts, and help strengthen natural resource management globally
Peningkatan Kemampuan Digital Masyarakat Melalui Pelatihan Website E-Commerce Berbasis AI Sarifah Agustiani; Aryanti, Riska; Wahyuni, Tri; Saepudin, Atang; Haliza Ramadhanti, Pristya; Roy Prasetya, Andreas
Darma Abdi Karya Vol. 4 No. 1 (2025): Darma Abdi Karya: Jurnal Pengabdian Kepada Masyarakat
Publisher : LPPM POLITEKNIK LP3I

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.38204/darmaabdikarya.v4i1.2338

Abstract

Kemampuan masyarakat dalam mengadopsi teknologi digital merupakan faktor krusial dalam mendukung pengembangan usaha lokal. Warga RT.010 Kelurahan Tegal Parang, Jakarta Selatan, yang mayoritas berprofesi sebagai pelaku usaha mikro seperti pengrajin, pengemudi ojek online, dan sopir taksi, masih menghadapi kendala signifikan dalam memanfaatkan teknologi digital, khususnya di bidang e-commerce dan kecerdasan buatan (AI). Rendahnya literasi digital dan keterbatasan akses terhadap pelatihan teknologi menjadi hambatan utama dalam memperluas jangkauan usaha mereka di era digital. Untuk mengatasi hal tersebut, dilakukan kegiatan pelatihan pembuatan website berbasis AI yang bertujuan meningkatkan kemampuan digital warga. Selain mengenalkan konsep dasar e-commerce dan AI, kegiatan ini membimbing peserta dalam membangun website usaha secara instan menggunakan platform ZipWP AI Website Builder tanpa memerlukan keterampilan pemrograman. Hasil pelatihan menunjukkan adanya peningkatan pemahaman dan keterampilan digital warga serta tumbuhnya semangat untuk mengelola usaha secara daring. Pelatihan ini juga berkontribusi dalam pembentukan ekosistem digital komunitas yang mendukung inklusi teknologi secara berkelanjutan. Dengan adanya kegiatan ini, diharapkan warga menjadi lebih siap menghadapi tantangan ekonomi digital sekaligus memperkuat ketahanan dan kemandirian komunitas dalam menghadapi era Revolusi Industri 4.0.
Comparative Optimization of EfficientNetB3, MobileNetV2, and ResNet50 for Waste Classification Agustiani, Sarifah; Haryani, Haryani; Junaidi, Agus; Putri, Rizky Rachma; Emiliana, Meutia Raissa
Jurnal Informatika Vol 12, No 2 (2025): October
Publisher : Universitas Bina Sarana Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31294/inf.v12i2.27533

Abstract

Waste management has become a critical challenge in efforts to maintain environmental sustainability and public health. Poorly managed waste can cause environmental pollution, reduce quality of life, and complicate recycling processes. To address this issue, this study aims to classify waste based on images while optimizing several deep learning architectures, namely EfficientNetB3, MobileNetV2, and ResNet50, to identify the best model for waste classification. The research methodology includes data collection, preprocessing, data augmentation, model development, and performance evaluation using accuracy, precision, recall, and F1-score metrics. The dataset, obtained from the Kaggle platform, consists of 4,650 images divided into six categories: battery, glass, metal, organic, paper, and plastic. The results show that EfficientNetB3 with the Adam optimizer achieved the best performance, with accuracy, precision, recall, and F1-score all at 93%, followed by ResNet50 at approximately 91%, and MobileNetV2 ranging from 70–73% depending on the optimizer. The use of different optimizers was found to influence model performance, and data augmentation helped improve generalization, especially for classes with limited samples. Limitations of this study include the relatively limited dataset coverage. Future research is recommended to expand the dataset and explore alternative or hybrid architectures. These findings demonstrate the potential of deep learning–based systems in supporting sustainable waste management.
Model Rapid Application Development (RAD) Untuk Perancangan Sistem Informasi Pengelolaan Surat (SIPERA) Pada Kelurahan Dalis, Sopiyan; Agustiani, Sarifah; Bahri, Syamsul; Wahyudin, Wahyudin; Prawikas, Amanda
IMTechno: Journal of Industrial Management and Technology Vol. 5 No. 1 (2024): Vol. 5 No. 1 (2024): Januari 2024
Publisher : LPPM Universitas Bina Sarana Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31294/imtechno.v5i1.2448

Abstract

Every government institution must have its own method of managing correspondence, either through a computerized system or still using a manual approach. In the context of managing incoming mail and outgoing mail in some institutions, there is the use of a manual approach that sometimes creates obstacles and constraints. One of the problems that often arises is the difficulty in finding data due to the excessive accumulation of letter archives. A similar situation is also encountered at the Muarasari Village Office, South Bogor Subdistrict, Bogor City, where the manual method used makes the process of searching for letters complicated and time-consuming. In order to overcome these problems, this research aims to design a computer-based incoming and outgoing mail management system to improve efficiency and effectiveness in managing these letters. In an effort to design a mail management system, the research approach involved interviews, observations, and using the Rapid Application Development (RAD) methodology, which emphasizes software development with relatively short time and the ability to adapt to changes. Through the implementation of this computerized mail management system, it is expected that the challenges in managing incoming and outgoing mail at the Muarasari Village Office can be overcome. Thus, this system is expected to increase efficiency and effectiveness in the management of letters, help improve the search process, and support overall improvement in the management of incoming and outgoing letters.
Penentuan Kelayakan Nasabah Untuk Kredit Motor Dengan Menggunakan Metode Topsis Hidayat, Rahmat; Agustiani, Sarifah
IJCIT (Indonesian Journal on Computer and Information Technology) Vol 8, No 2 (2023): IJCIT November 2023
Publisher : LPPM Universitas Bina Sarana Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31294/ijcit.v8i2.15390

Abstract

Transportasi merupakan salah satu kebutuhan dalam kehidupan sehari-hari yang sering digunakan hampir oleh setiap kalangan. Salah satu transportasi yang paling sering digunakan saat ini adalah kendaraan bermotor atau sepeda motor. Hal ini didukung dengan harga motor yang relatif masih terjangkau serta adanya sistem kredit kepemilikan motor sehingga menjadi peluang dalam memudahkan masyarakat untuk memilikinya. Pada sistem kredit kepemilikan motor, setiap penyedia jasa memiliki kriteria tertentu dalam menentukan siapa yang layak diberikan jasa kredit motor ini. Hal ini untuk meminimalisir  terjadinya kerugian pada salah satu pihak. Penelitian ini bertujuan untuk menentukan kelayakan nasabah untuk kredit motor menggunakan metode topsis. Metode topsis memiliki konsep yang sederhana, mudah dipahami, serta mampu mengukur kinerja relatif dari alternatif keputusan dalam bentuk matematis yang sederhana. Berdasarkan penelitian yang telah dilakukan dapat disimpulakan bahwa dari 6 nasabah yang menjadi sample dalam penilitian terdapat 2 nasabah layak, 1 nasabah layak dipertimbangkan dan 3 nasabah lainnya tidak layak. Dengan demikian metode topsis ini dapat digunakan untuk mempermudah dalam proses pengambilan keputusan untuk menentukan kelayakan nasabah kredit motor. Transportation is a necessity in daily life that is often used by almost every group. One of the most frequently used forms of transportation today is motorized vehicles or motorbikes. This is supported by the relatively affordable price of motorbikes and the existence of a credit system for motorbike ownership, making it an opportunity to make it easier for people to own them. In the motorbike ownership credit system, each service provider has certain criteria in determining who is eligible to be provided with this motorbike credit service. This is to minimize losses to one party. This research aims to determine customer eligibility for motorbike credit using the topsis method. The topsis method has a simple concept, is easy to understand, and is able to measure the relative performance of decision alternatives in simple mathematical form. Based on the research that has been carried out, it can be concluded that of the 6 customers sampled in the research, there are 2 worthy customers, 1 customer is worthy of consideration and 3 other customers are not worthy. Thus, this topsis method can be used to simplify the decision-making process to determine the suitability of motorbike credit customers.
Waste Classification using EfficientNetB3-Based Deep Learning for Supporting Sustainable Waste Management Agustiani, Sarifah; Junaidi, Agus; Aryanti, Riska; Kamil, Anton Abdul Basah
Bulletin of Informatics and Data Science Vol 4, No 1 (2025): May 2025
Publisher : PDSI

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61944/bids.v4i1.108

Abstract

Waste management is a critical issue in sustainable development, particularly in large urban areas that generate a high volume of waste daily. One of the main challenges is the absence of a fast, accurate, and efficient waste sorting system. This study aims to develop a waste classification model using deep learning based on the EfficientNetB3 architecture to support more sustainable waste management. The model was trained on a dataset obtained from a Kaggle repository, consisting of 4,650 images evenly distributed across six waste categories: batteries, glass, metal, organic, paper, and plastic (775 images per class). The training and evaluation were conducted using a supervised image classification approach. The model achieved an overall accuracy of 93%, with average precision, recall, and F1-score values of 93%. Among all categories, organic waste achieved the highest F1-score (0.99), followed by paper (0.97) and batteries (0.97), while plastic and metal categories obtained F1-scores of 0.89. These results demonstrate that the EfficientNetB3 architecture is effective in performing multi-class waste classification. This model has the potential to be implemented in camera-based waste sorting systems such as smart bins or automated recycling units, thereby contributing to the reduction of unprocessed waste and supporting the achievement of Sustainable Development Goal (SDG) 12: responsible consumption and production
Chicken Disease Detection Based on Fases Image Using EfficientNetV2L Model Mustopa, Ali; Sasongko, Agung; Nawawi, Hendri Mahmud; Wildah, Siti Khotimatul; Agustiani, Sarifah
Sistemasi: Jurnal Sistem Informasi Vol 12, No 3 (2023): Sistemasi: Jurnal Sistem Informasi
Publisher : Program Studi Sistem Informasi Fakultas Teknik dan Ilmu Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32520/stmsi.v12i3.2807

Abstract

Livestock farming requires technological innovation to increase productivity and efficiency. Chickens are a livestock animal with good market prospects. However, not all farmers understand about chicken diseases and signs of sickness. Detection of chicken diseases can be done through various methods, one of which is by looking at the shape of the chicken's feces. Images in feces can be detected using machine learning. Convolutional Neural Networks (CNN) are used to speed up disease prediction. Transfer learning is used to leverage knowledge that has been learned by previous models. In this study, we propose our own CNN architecture model and present research by building a new model to detect and classify diseases in chickens through their feces. The model training process is carried out by inputting training data and validation data, the number of epochs, and the created checkpointer object. The hyperparameter tuning stage is carried out to increase the accuracy rate of the model. The research is conducted by testing datasets obtained from the Kaggle repository which has images of coccidiosis, salmonella, Newcastle, and healthy feces. The results of the study show that our proposed model only achieves an accuracy rate of 93%, while the best accuracy rate in the study is achieved by using the EfficientNerV2L model with the RMSProp optimizer, which is 97%.
Pengembangan Sistem Informasi Akademik untuk Meningkatkan Efektivitas Pengelolaan Data pada SMK Mihadunal Ula Agustiani, Sarifah; Pribadi, Denny; Dalis, Sopiyan; Wildah, Siti Khotimatul; Mustopa, Ali
Reputasi: Jurnal Rekayasa Perangkat Lunak Vol. 4 No. 1 (2023): Mei 2023
Publisher : LPPM Universitas Bina Sarana Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31294/reputasi.v4i1.1992

Abstract

Teknologi informasi memiliki peran penting dalam mendukung efisiensi dan efektivitas pengelolaan data di lembaga pendidikan. SMK Mihadunal Ula, sebagai sekolah menengah kejuruan di Kabupaten Sukabumi, menghadapi tantangan dalam pengelolaan data akademik yang masih dilakukan secara manual. Hal ini menyebabkan berbagai masalah seperti kesalahan data, kesulitan akses informasi, dan keterlambatan dalam pengolahan data. Penelitian ini bertujuan untuk mengembangkan Sistem Informasi Akademik yang dapat meningkatkan efektivitas pengelolaan data pada SMK Mihadunal Ula. Metode pengembangan yang digunakan adalah pengembangan sistem Rapid Application Development (RAD) yang melibatkan proses analisis, desain, implementasi, dan evaluasi. Melalui pengembangan sistem informasi akademik, diharapkan pengelolaan data di SMK Mihadunal Ula dapat lebih terintegrasi, akurat, dan mudah diakses. Sistem ini akan menyediakan fitur-fitur penting seperti pendaftaran siswa, penjadwalan, dan pembayaran yang dapat diakses oleh siswa, guru, dan staf administrasi. Dengan adanya sistem informasi yang handal, diharapkan efisiensi operasional sekolah dapat ditingkatkan, kesalahan manusia dapat diminimalisir, dan pengambilan keputusan dapat lebih baik. Hasil penelitian ini menunjukkan bahwa implementasi Sistem Informasi Akadmik pada SMK Mihadunal Ula memberikan manfaat yang signifikan. Siswa dapat dengan mudah mendaftar, memperoleh informasi jadwal pelajaran, dan melakukan pembayaran secara efisien. Guru dan staf administrasi juga mendapatkan kemudahan dalam pengolahan data dan mengakses informasi yang diperlukan. Selain itu, penggunaan sistem informasi ini diharapkan dapat meningkatkan citra dan reputasi SMK Mihadunal Ula sebagai lembaga pendidikan yang modern dan berkualitas
Transformasi Digital untuk Mewujudkan Ruang Publik Yang Lebih Cerdas Haryani, Haryani; Agustiani, Sarifah; Junaidi, Agus; Wahyudin, Wahyudin; Agus Junaidi
Jurnal Abdimas Ekonomi dan Bisnis Vol. 4 No. 2 (2024): Jurnal Abdimas Ekonomi dan Bisnis
Publisher : LPPM Universitas Bina Sarana Informatika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31294/abdiekbis.v4i2.7289

Abstract

The purpose of this activity is to provide solutions to problems in RPTRA in the form of RPTRA management information systems and provide information technology training in the form of RPTRA website management. The problem in the management of RPTRA so far is that there has been no integrated data, both data on guest visits, activity agendas, and events carried out by RPTRA and external parties that use the Annur RPTRA land. The main problems related to digital infrastructure at RPTRA Annur are the availability of access to digital information such as the use of devices for digital learning for children, access to information and online registration of activities, communication and coordination between RPTRA managers and the community. The expected result in Community Service at RPTRA Annur is to make a product in the form of an RPTRA website profile that contains an agenda of activities and events, and a guest visit book that uses land at RPTRA. In addition, with this digital transformation activity, RPTRA managers also gain increased knowledge in using the Canva application for the learning process carried out.