Claim Missing Document
Check
Articles

Found 3 Documents
Search

Deteksi Jamur Beracun dengan Algoritma Convolutional Neural Network dan Arsitektur EfficientNet-B0 Mauludy, Muhammad Wildan; Rulyana, Devita; Hardjianto, Mardi
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 8, No 1 (2024): Januari 2024
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v8i1.7276

Abstract

Indonesia is a tropical country that has abundant natural resources and biodiversity, one of which is mushrooms. Mushrooms have various shapes and types. Some of them contain mushrooms that cannot be consumed because they contain toxins that will have an impact on human health. Mushroom species that can be consumed sometimes have a similar shape to mushrooms that cannot be consumed, causing cases of poisoning due to consuming the wrong mushrooms. This research focuses on detecting poisonous mushrooms using a Convolutional Neural Network (CNN) with the EfficientNet-B0 architecture. Mushroom data was obtained from Kaggle, and after praprocessing, the model was trained by varying the number of epochs and batch size. Based on the results of research and discussion on the detection of poisonous and non-toxic mushrooms, it is concluded that the CNN algorithm with the EfficientNet-B0 architecture can differentiate between poisonous and non-toxic mushrooms with a high level of accuracy. In scenario testing, the model trained using batch size 32 had an accuracy of 84.2% and loss of 0.39, precision of 0.855, recall of 0.805, and f1 score of 0.815. This shows that the CNN architecture EfficientNet-B0 is an efficient and accurate approach in classifying poisonous and non-poisonous mushrooms. Apart from that, this research also found that parameters such as the number of epochs and the number of batch sizes influence the model training process.
MEATCHECK: DETEKSI KUALITAS DAGING SAPI BERBASIS MOBILE DEEP LEARNING Muh. Wildan Mauludy; Goenawan Brotosaputro; Mardi Hardjianto
Jurnal Ilmiah Informatika Komputer Vol 30, No 1 (2025)
Publisher : Universitas Gunadarma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35760/ik.2025.v30i1.14265

Abstract

Beef is an important foodstuff that affects consumer satisfaction and market value in the meat industry. The purpose of this research is to develop a model to classify beef quality using the transfer learning method. The data collection method is carried out through taking pictures of beef, which are then labeled based on their quality. Classification uses a transfer learning architecture that can improve the performance of the machine learning model generated for the classification of fresh and rotten meat. The model was tested by looking at accuracy, precision, recall, and f1-score. The results showed an accuracy of 61%, precision of 60.78%, recall of 61%, and an f1-score of 60.89%, which was achieved with a learning rate of 0.1, 10 epoch, and batch size of 8. Conclusion, the model developed with the transfer learning algorithm MobileNetV2 was able to classify the quality of beef with a good level of accuracy. The prototype of the developed system can provide real-time predictions, help consumers choose quality meat, and increase market value. Next, it is recommended to increase accuracy and develop models by increasing the size of the dataset and exploring other, more complex architectures.
Pengembangan Sistem Informasi Pengarsipan File Menggunakan Metode R&D Mauludy, Muh Wildan; Fathahillah
Journal of Informatics Management and Information Technology Vol. 5 No. 2 (2025): April 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/jimat.v5i2.484

Abstract

This research adopts the Research and Development method technique or better known as R&D, which is developing a product in the form of a system specifically designed to archive files in the form of documents at PT. THE AGE OF SELF-RELIANCE. The approach used is the ADDIE (Analyze, Design, Develop, Implement, and Evaluate) model. The subject of the study was an employee who used a system with samples taken from 2 employees through a simple random sampling technique, the limitations of the employees made the samples taken few. Data was collected using questionnaires and in-depth interviews, then analyzed based on ISO 25010 criteria which included eight characteristics: functionality, reliability, performance efficiency, portability, maintainability, usability, compatibility, and security. The test results show that the characteristics of functional suitability reach 100%, reliability 97%, performance efficiency 94%, portability 100%, usability 90%, maintainability 100%, security with alert level 0, and compatibility 100%. The conclusion of the study is that this system can be used effectively in the archiving of documents related to corporate archives and has met all the characteristics of ISO 25010. It is recommended that companies conduct periodic evaluations and updates of the system to ensure optimal performance and better data security.