Article Per Year (5 Year)
p-Index From 2020 - 2025
0.408
P-Index
This Author published in this journals
All Journal Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Jurnal INFOTEL
Aryanto, Andreas Sahir
Unknown Affiliation
Computer Science & IT Electrical & Electronics Engineering

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Addressing Overfitting in Dermatological Image Analysis with Bayesian Convolutional Neural Network Zulfa, Mulki Indana; Aryanto, Andreas Sahir; Wijonarko, Bintang Abelian; Ahmed, Waleed Ali
Jurnal Ilmiah Teknik Elektro Komputer dan Informatika Vol. 10 No. 2 (2024): June
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26555/jiteki.v10i2.29177

Abstract

VGG, ResNet, and DenseNet are popular convolutional neural network (CNN) designs for transfer learning (TL), aiding dermatological image processing, particularly in skin cancer categorization. These TL-CNN models build extensive neural network layers for effective image classification. However, their numerous layers can cause overfitting and demand substantial computational resources. The Bayesian CNN (BCNN) technique addresses TL-CNN overfitting by introducing uncertainty in model weights and predictions. Research contributions are (i) comparing BCNN with three TL-CNN architectures in dermatological image processing and (ii) examining BCNN ability to mitigate overfitting through weight perturbation and uncertainty during training. BCNN uses flipout layers to perturb weights during training, guided by the KL divergence and Binary Cross Entropy (BCE) loss function. The dataset used is the ISIC Challenge 2017, categorized as malignant and benign skin tumors. The simulation results show that three TL-CNN architectures, namely VGG-19, ResNet-101, and DenseNet-201, obtained training accuracies of 96.65%, 100%, and 97.70%, respectively. However, all three were only able to achieve a maximum validation accuracy of around 78%. In contrast, BCNN can produce training and validation accuracy of 81.30% and 80%, respectively. The difference in training and validation accuracy values produced by BCNN is only 1.3%. Meanwhile, the three TL-CNN architectures are trapped in an overfitting condition with a difference in training and validation values of around 20%. Therefore, BCNN is more reliable for dermatological image processing, especially for skin cancer images.
Model Siklus Waktu Lampu Lalu Lintas Cerdas Menggunakan Fuzzy Mamdani Zulfa, Mulki Indana; Aryanto, Andreas Sahir; Fadli, Ari
JURNAL INFOTEL Vol 16 No 2 (2024): May 2024
Publisher : LPPM INSTITUT TEKNOLOGI TELKOM PURWOKERTO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20895/infotel.v16i2.1106

Abstract

The growth of motorized vehicles in Indonesia has increased significantly. According to data from the Central Bureau of Statistics, the number of motorized vehicles in Indonesia has increased by around 10% each year in the last five years. One of the negative impacts of the increasing number of motorized vehicles is traffic congestion. Traffic congestion has become a serious problem in several cities in Indonesia. One of the causes is the increase in the number of vehicles at road intersections, which has an impact on congestion and the safety of road users. The rapid growth in the number of vehicles requires a more comprehensive strategy to reduce congestion and accidents at road intersections. Therefore, the need for Intelligent Transportation System, especially on the time-cycle configuration of intelligent red light is very important. This research aims to model the time-cycle of the red light using the Mamdani Fuzzy Inference System to simulate the green light time configuration so as to reduce the waiting time of road users at highway intersections. The simulation results show that the time-cycle configuration and green light time length of the Mamdani Fuzzy calculation are more varied relative to the number of vehicles. The values are relatively smaller than 6 to 54 seconds from the time configuration set by the local Department of Transportation. This shows a time efficiency for road users of up to 27%, which means that road users can complete trips 6 to 13 seconds faster.
Co-Authors Ahmed, Waleed Ali Ari Fadli Mulki Indana Zulfa, Mulki Indana Wijonarko, Bintang Abelian