Article Per Year (5 Year)
p-Index From 2020 - 2025
1.556
P-Index
This Author published in this journals
All Journal Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Telematika Indonesian Journal on Computing (Indo-JC) Jurnal Sisfokom (Sistem Informasi dan Komputer) JIPI (Jurnal Ilmiah Penelitian dan Pembelajaran Informatika) Journal of Dinda : Data Science, Information Technology, and Data Analytics Jurnal INFOTEL
Ikke Dian Oktaviani
Telkom University
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Education Electrical & Electronics Engineering

Published : 7 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 7 Documents
Search

 1 
Imputasi KNN terhadap Nilai yang Hilang dari Prediksi Durasi Hujan Berbasis Regresi pada Data BMKG Ikke Dian Oktaviani; Aji Gautama Putrada
JURNAL INFOTEL Vol 14 No 4 (2022): November 2022
Publisher : LPPM INSTITUT TEKNOLOGI TELKOM PURWOKERTO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20895/infotel.v14i4.840

Abstract

The prediction of rain duration based on data from the Meteorology, Climatology, and Geophysics Agency (BMKG) is an important issue but remains an open problem. At the same time, several studies have shown that missing values can cause a decrease in the performance of the model in making predictions. This study proposes k-nearest neighbors (KNN) imputation to overcome the problem of missing values in predicting rain duration. The source of the rain duration prediction dataset is the BMKG data. We compared gradient boosting regression (GBR), adaptive boosting regression (ABR), and linear regression (LR) for the regression model for predicting rain duration. We compared the KNN imputation method with several benchmark methods, including zero imputation, mean imputation, and iterative imputation. Parameters r2, mean squared error (MSE) and mean bias error (MBE) measure the performance of these imputation methods. The test results show that for rain duration prediction using the regression method, GBR shows the best performance, both for train data and test data with r2 = 0.915 and 0.776, respectively. Then our proposed KNN imputation has the best performance for missing value imputation compared to the benchmark imputation method. The prediction values of r2 and MSE when using KNN imputation at Missing Percentage = 90% are 0.71 and 0.36, respectively.
QUIDS: A Novel Edge-Based Botnet Detection with Quantization for IoT Device Pairing Aji Gautama Putrada; Nur Alamsyah; Mohamad Nurkamal Fauzan; Sidik Prabowo; Ikke Dian Oktaviani
Indonesia Journal on Computing (Indo-JC) Vol. 8 No. 3 (2023): December 2023
Publisher : School of Computing, Telkom University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34818/INDOJC.2023.8.3.878

Abstract

Advanced machine learning has managed to detect IoT botnets. However, conflicts arise due to complex models and limited device resources. Our research aim is on a quantized intrusion detection system (QUIDS), an edge-based botnet detection for IoT device pairing. Using knearest neighbor (KNN) within QUIDS, we incorporate quantization, random sampling (RS), and feature selection (FS). Initially, we simulated a botnet attack, devised countermeasures via a sequence diagram, and then utilized a Kaggle botnet attack dataset. Our novel approach includes RS, FS, and 16-bit quantization, optimizing each step empirically. The test results show that employing a mean decrease in impurity (MDI) by FS reduces features from 115 to 30. Despite a slight accuracy drop in KNN due to RS, FS, and quantization sustain performance. Testing our model revealed 1200 RS samples as optimal, maintaining performance while reducing features. Quantization to 16-bit doesn’t alter feature value distribution. Implementing QUIDS increased the compression ratio (CR) to 175×, surpassing RS+FS threefold and RS by 13 times. This novel method emerges as the most efficient in CR.
CNN Pruning for Edge Computing-Based Corn Disease Detection with a Novel NG-Mean Accuracy Loss Optimization Putrada, Aji Gautama; Oktaviani, Ikke Dian; Fauzan, Mohamad Nurkamal; Alamsyah, Nur
Telematika Vol 17, No 2: August (2024)
Publisher : Universitas Amikom Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35671/telematika.v17i2.2899

Abstract

Plant disease detection studies disease attacks in plants detected on the leaves using computer vision. However, some plant disease detection solutions still utilize cloud computing, where the problems include slow processing times and misuse of data privacy. This study aims to evaluate the performance of convolutional neural network (CNN) pruning in edge computing-based plant disease detection. We use Kaggle's plant disease image dataset, which contains three corn diseases. We also created an edge computing system architecture for plant disease detection utilizing the latest communication technology and middleware. Next, we developed an optimal CNN model for plant disease detection using grid search. We pruned the CNN model in the final step and tested its performance. In this step, we developed a novel normalized-geometric mean (NG-mean) method for accuracy loss optimization. The test results show that class weights can optimize specificity and g-mean on the imbalanced dataset, with values of 0.995 and 0.983, respectively. The grid search results then optimize the optimization method's hyperparameters, learning rate, batch size, and epoch to achieve the highest accuracy of 0.947. Applying pruning produces several models with variations in sparsity and scheduling methods. We used the new NG-mean method to find the best compressed model. It had constant scheduling, 0.8 sparsity, a mean accuracy loss of 1.05%, and a CR of 2.71×. This study enhances the efficiency and privacy of plant disease detection by utilizing edge computing and optimizing CNN models, leading to faster processing and better data security. Future work could explore the application of the novel NG-Mean method in other domains and the integration of additional plant species and diseases into the detection system.
ANALYSIS OF THE QUALITY OF THE INFINITE LEARNING TRAINING COURSE WEBSITE AT INFINITE LEARNING USING THE WEBQUAL 4.0 METHOD AND IMPORTANCE PERFORMANCE ANALYSIS (IPA) Avicena, Ginanjar Ilmi Nur; Makky, Muhammad Al; Oktaviani, Ikke Dian
JIPI (Jurnal Ilmiah Penelitian dan Pembelajaran Informatika) Vol 10, No 1 (2025)
Publisher : STKIP PGRI Tulungagung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29100/jipi.v10i1.6067

Abstract

The training course sector in Indonesia has experienced rapid development in recent years. Many institutions offer various training courses to enhance skills and knowledge across different fields, from information technology to human resource management, with one such institution being Infinite Learning. The objective of this research is to assess the quality of the website and identify the factors that can improve its quality. Utilizing the Webqual 4.0 method as an evaluation indicator and Importance Performance Analysis to categorize the factors into quadrants based on the analysis of Importance Performance Analysis. Before data grouping, validity tests, reliability tests, and gap analysis between user expectations and perceptions need to be conducted. This step aims to strengthen the conclusions and recommendations produced after the research is completed. After conducting the research, the final results show that the Infinite Learning training course website has not met user expectations. This is evident from the gap analysis, which resulted in a score of -0.63, indicating that user expectations or importance are higher than the performance of the Infinite Learning training course website. The results of this research show that all performance values of each attribute are greater than the Importance values, indicating that users are satisfied with the services of the Infinite Learning training course website, with the highest value in the dimension of interaction quality at 0.93.
CNN-LSTM for MFCC-based Speech Recognition on Smart Mirrors for Edge Computing Command Aji Gautama Putrada; Ikke Dian Oktaviani; Mohamad Nurkamal Fauzan; Nur Alamsyah
Indonesian Journal of Data Science, IoT, Machine Learning and Informatics Vol 4 No 2 (2024): August
Publisher : Research Group of Data Engineering, Faculty of Informatics

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20895/dinda.v4i2.1504

Abstract

Smart mirrors are conventional mirrors that are augmented with embedded system capabilities to provide comfort and sophistication for users, including introducing the speech command function. However, existing research still applies the Google Speech API, which utilizes the cloud and provides sub-optimal processing time. Our research aim is to design speech recognition using Mel-frequency cepstral coefficients (MFCC) and convolutional neural network–long short-term memory (CNN-LSTM) to be applied to smart mirror edge devices for optimum processing time. Our first step was to download a synthetic speech recognition dataset consisting of waveform audio files (WAVs) from Kaggle, which included the utterances “left,” “right,” “yes,” “no,” “on,” and “off. ” We then designed speech recognition by involving Fourier transformation and low-pass filtering. We benchmark MFCC with linear predictive coding (LPC) because both are feature extraction methods on speech datasets. Then, we benchmarked CNN-LSTM with LSTM, simple recurrent neural network (RNN), and gated recurrent unit (GRU). Finally, we designed a smart mirror system complete with GUI and functions. The test results show that CNN-LSTM performs better than the three other methods with accuracy, precision, recall, and an f1-score of 0.92. The speech command with the best precision is "no," with a value of 0.940. Meanwhile, the command with the best recall is "off," with a value of 0.963. On the other hand, the speech command with the worst precision and recall is "other," with a value of 0.839. The contribution of this research is a smart mirror whose speech commands are carried out on the edge device with CNN-LSTM.
Implementation of Doodle Jump Game Based on Accelerometer Sensor and Kalman Filter Danuwijaya, Edghar; Putra, Yohanes Armenian; Nuha, Hilal Hudan; Oktaviani, Ikke Dian
Jurnal Sisfokom (Sistem Informasi dan Komputer) Vol. 14 No. 1 (2025): JANUARY
Publisher : ISB Atma Luhur

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32736/sisfokom.v14i1.2259

Abstract

The doodle jump game is a video game with a jumping game model assisted by accelerometer sensor technology. Placing the accelerometer sensor in the doodle jump game is a very appropriate solution to determine the accuracy of the values on the sensor. The accelerometer sensor can be measured in real time, however applying a small force to the sensor can result in interference with measurement accuracy. Therefore, creating the measurement results you need using filters can help reduce noise. The method used to use this filter is the Kalman Filter algorithm. The use of the Kalman Filter method can provide a stable level of accuracy in the movements of the main characters in the game and the accelerometer sensor so that it can become a precise algorithm. Apart from that, the use of the Kalman Filter as a tool or method for measuring numbers to provide a solution to improve the design of the previous developer.
A Hybrid Genetic Algorithm-Random Forest Regression Method for Optimum Driver Selection in Online Food Delivery Putrada, Aji Gautama; Alamsyah, Nur; Oktaviani, Ikke Dian; Fauzan, Mohamad Nurkamal
Jurnal Ilmiah Teknik Elektro Komputer dan Informatika Vol. 9 No. 4 (2023): December
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26555/jiteki.v9i4.27014

Abstract

The online food delivery trend has become rapid due to the COVID-19 incident, which limited mobility, while the broader challenge in the online food delivery system is maximizing quality of service (QoS). However, studies show that driver selection and delivery time are important in customer satisfaction. The solution is our research aim, which is the selection of optimal drivers for online food delivery using random forest regression and the genetic algorithm (GA) method. Our research contribution is a novel approach to minimizing delivery time in online food delivery by combining a random forest regression model and genetic algorithms. We compare random forest regression with three other state-of-the-art regression models: linear regression, k-nearest neighbor (KNN), and adaptive boosting (AdaBoost) regression. We compare the four models with metrics including , mean squared error (MSE), root mean squared error (RMSE), mean total error (MAE), and mean absolute percentage error (MAPE). We use the optimum model as the fitness function in GA. The test results show that random forest performs better than linear, KNN, and AdaBoost regression, with an , RMSE, and MAE value of 0.98, 54.3, and 11, respectively. We leverage the optimum random forest regression model as the GA fitness function. The best efficiency is reducing the delivery time from 54 to 15 minutes, achieved through rigorous testing on various cases. In addition, by completing this research, we also achieve some practical implications, such as an increase in customer satisfaction, a reduction in cost, and a paramount finding in the field of data-driven decision-making. The first key finding is an optimum driver selection model in random forest regression, while the second is an optimum driver selection model in GA.
Co-Authors Aji Gautama Putrada Avicena, Ginanjar Ilmi Nur Danuwijaya, Edghar Fauzan, Mohamad Nurkamal Fauzan, Mohamad Nurkamal Mohamad Nurkamal Fauzan Mohamad Nurkamal Fauzan Muhammad Al Makky Nuha, Hilal H Nur Alamsyah Nur Alamsyah, Nur Putra, Yohanes Armenian Sidik Prabowo