Jurnal Nasional Pendidikan Teknik Informatika : JANAPATI

The Implementation of Bayesian Optimization for Automatic Parameter Selection in Convolutional Neural Network for Lung Nodule Classification

2024-08-05T23:46:09+00:00

Lung cancer's high mortality rate makes early detection crucial. Machine learning techniques, especially convolutional neural networks (CNN), play a very important role in lung nodule detection. Traditional machine learning approaches often require manual feature extraction, while CNNs, as a specialized type of neural network, automatically learn features directly from the data. However, tuning CNN hyperparameters, such as network structure and training parameters, is computationally intensive. Bayesian Optimization offers a solution by efficiently selecting parameter values. This study develops a CNN classification model with hyperparameter tuning using Bayesian Optimization, achieving a 97.2% accuracy. Comparatively, Particle Swarm Optimization and Genetic Algorithm methods each resulted in 96.4% accuracy. The research concludes that Bayesian Optimization is an effective approach for CNN hyperparameter tuning in lung nodule classification.

Deep Learning for Karolinska Sleepiness Scale Classification Based On Eye Aspect Ratio with SMOTE-Enhanced Data Balancing

2024-10-15T02:07:36+00:00

This paper addresses the challenge of accurately classifying sleepiness levels based on the Karolinska Sleepiness Scale (KSS) using Eye Aspect Ratio (EAR) data, especially when class imbalance leads to biased predictions. The research proposes a deep learning framework that integrates a Multi-Layer Perceptron (MLP) with the Synthetic Minority Over-sampling Technique (SMOTE) to balance the training data. EAR features, representing eye closure patterns, are extracted from video frames, and SMOTE is applied to generate synthetic data for underrepresented sleepiness classes. By training the MLP model on this balanced dataset, the system achieves a 97.6% classification accuracy in distinguishing four distinct sleepiness levels based on the KSS, demonstrating its effectiveness in reducing prediction bias and managing class imbalance, both crucial for real-time drowsiness detection systems

The Implementation of Enterprise Resource Planning During the Product Design Process Through the Process of Design Thinking

2024-07-25T10:43:39+00:00

The implementation of Enterprise Resource Planning (ERP) systems in the product design phase plays a crucial role in modern industries. The product design phase with the design thinking approach produces an innovative product that meets user requirements. The product design process, which begins with capturing user requirements and culminating in a thorough specification of a finished product, requires extensive data and expertise. Iterative product design processes contribute to the complexity of the data that must be managed within an organization. This study involved the implementation of an Enterprise Resource Planning (ERP) software named Odoo within the product design process using a design thinking methodology. By examining the student practicum activities in automotive design, starting from the user survey phase and going all the way to the component design details, a comprehensive ERP system was developed. This system is capable of seamlessly integrating all the data throughout the entire process. Based on the outcomes of testing and assessment, it can be concluded that the modules in Odoo software can be effectively integrated into the product design process. Effective processes in integrating ERP into product design phases can improve production quality and efficiency as well as facilitate greater flexibility and innovation. Implementing ERP throughout the product design phase results in a seamless flow of information, enhanced inventory control, and overall productivity enhancement. This ultimately leads to operational efficiency, competitive advantage, and high user satisfaction in the industry.

Detection of UDP Flooding DDoS Attacks on IoT Networks Using Recurrent Neural Network

2024-10-15T03:44:11+00:00

Internet of Thing (IoT) is a concept where an object can transfer data through a network without requiring human interaction. Complex IoT networks make it vulnerable to cyber attacks such as DDoS UDP Flood attacks, UDP Flood attacks can disrupt IoT devices. Therefore, this study proposes an attack detection method using a deep learning approach with the Recurrent Neural Network (RNN) method. This study uses Principle Component Analysis (PCA) to reduce the feature dimension, before learning using RNN. The purpose of this study is to test the combined performance of the PCA and RNN methods to detect DDoS UDP Flood attacks on IoT networks. The testing in this study used 10 datasets sourced from CICIOT2023 containing UDP Flood and Benign DDoS traffic data, and the testing was carried out using three epoch parameters (iterations), namely 10, 50, and 100. The test results using RNN epoch 100 were superior, showing satisfactory performance with an accuracy value of 98%, precision of 99%, recall of 99%, and f1-score of 99%. Based on the experimental results, it can be concluded that combining PCA and RNN is able to detect UDP Flooding attacks by showing high accuracy.

Network Intrusion Detection Using Transformer Models and Natural Language Processing for Enhanced Web Application Attack Detection

2024-10-15T06:39:36+00:00

The increasing frequency and complexity of web application attacks necessitate more advanced detection methods. This research explores integrating Transformer models and Natural Language Processing (NLP) techniques to enhance network intrusion detection systems (NIDS). Traditional NIDS often rely on predefined signatures and rules, limiting their effectiveness against new attacks. By leveraging the Transformer's ability to capture long-term dependencies and the contextual richness of NLP, this study aims to develop a more adaptive and intelligent intrusion detection framework. Utilizing the CSIC 2010 dataset, comprehensive preprocessing steps such as tokenization, stemming, lemmatization, and normalization were applied. Techniques like Word2Vec, BERT, and TF-IDF were used for text representation, followed by the application of the Transformer architecture. Performance evaluation using accuracy, precision, recall, F1 score, and AUC demonstrated the superiority of the Transformer-NLP model over traditional machine learning methods. Statistical validation through Friedman and T-tests confirmed the model's robustness and practical significance. Despite promising results, limitations include the dataset's scope, computational complexity, and the need for further research to generalize the model to other types of network attacks. This study indicates significant improvements in detecting complex web application attacks, reducing false positives, and enhancing overall security, making it a viable solution for addressing increasingly sophisticated cybersecurity threats

Classification of Lung Diseases in X-Ray Images Using Transformer-Based Deep Learning Models

2024-07-06T05:04:18+00:00

This research evaluates the performance of two Transformer models, the Vision Transformer (ViT) and Swin Transformer, in the analysis of thoracic X-ray images. The study's objective is to determine whether Transformer models can enhance diagnostic accuracy for lung diseases, considering challenges such as early symptom variability and similar radiological signs. The dataset includes 21,165 X-ray images, featuring 3,616 COVID-19 cases, 10,192 normal images, 6,012 images of Lung Opacity, and 1,345 pneumonia images. Model development involved tuning hyperparameters such as epoch numbers and optimizer choice. The results indicate that using the AdamW and Adamax optimizers achieves an optimal balance between computational efficiency and accuracy. The Swin Transformer model, using the Adamax optimizer, reached the highest testing accuracy of 96.10% in 33,802.70 seconds, while the Vision Transformer achieved a testing accuracy of 95.10% in 33,503.10 seconds.

Data Mining Analysis of Moodle Learning Data and Student Perceptions During and After the Covid-19 Pandemic

2024-09-25T22:44:33+00:00

This study examines the academic performance of students from the 2020 and 2023 cohorts, highlighting differences in activity, attendance, task completion, midterm and final exam scores, and perceptions of educational metrics. A data mining approach was applied to predict students' GPA using Decision Tree, Random Forest, Multinomial Naïve Bayes, and Gaussian Naïve Bayes algorithms. The Gaussian Naïve Bayes model showed the highest accuracy of 0.93 for the 2020 cohort and 0.92 for the 2023 cohort, with the lowest error rate making it the most effective predictor. Feature importance analysis revealed that task completion and exam scores were the most influential factors, while students' perceptions had a lesser impact. The findings suggest that direct academic metrics should be the focus for improving student performance. This study emphasizes the need for further refinement of predictive models and suggests incorporating both academic metrics and student perceptions for a holistic understanding of student performance.

The Data-Driven Approach in Transitioning Organizational Strategies and Capabilities: Insights from Indonesia's National Narcotics Agency

2024-09-11T07:55:54+00:00

Anti-narcotics prevention measures, such as urine sampling of suspect offenders, citizen reporting of suspect narcotic activities, public education, or legal consultations used to be performed at provincial and city levels. To improve effectiveness and efficiency, Indonesia’s National Narcotics Agency (Badan Narkotika Nasional or BNN) centralized such initiatives by introducing BOSS (BNN One-Stop Service), an integrated service information system provided to the public. However, at present data generated by BOSS has not been fully exploited in the design of BNN strategy. The objective of this study is to explore the untapped potential of BOSS data to improve BNN strategy and capabilities, focusing on preventing and eradicating narcotics abuse. The methodology used is descriptive qualitative, with data collection through document analysis and interviews. This study is expected to provide a preliminary interpretation of how BOSS data can improve BNN's ability to fight narcotics abuse more effectively and efficiently. The results of the study show that the integration of BOSS data can significantly optimize the efficiency, analytical capabilities, and responsiveness of BNN in dealing with narcotics abuse, showing that the use of strategic data from BOSS is the key to BNN's digital transformation for a more effective narcotics prevention and eradication strategy.

Optimizing Healthcare Performance Through Electronic Medical Records: An Efficiency Analysis

2024-11-07T05:31:55+00:00

Burnout among healthcare workers has become a significant issue in the medical field, partially due to the adaptation process to Electronic Medical Records (EMR). While EMR technology is designed to enhance efficiency and accuracy in patient care, it often poses challenges during implementation. This study aims to examine the impact of medical record digitalization on healthcare worker performance, mediated by high-quality data access and data-driven decision-making at Wangaya Regional Hospital. Using the SEM-PLS method and involving 244 healthcare workers, the research reveals that medical record digitalization significantly improves data access and data-driven decision-making. The findings indicate that EMR plays a crucial role in enhancing healthcare worker performance by facilitating quicker, more accurate, and up-to-date information access, ultimately improving service efficiency and effectiveness. These results support the implementation of digital transformation in medical record management to improve healthcare worker performance and, consequently, the overall quality of healthcare services

Facial Expression Detection System for Students in Classroom Learning Process Using YoloV7

2024-08-12T15:16:49+00:00

The utilization of technology in education is not only about using hardware or software, but also how technology can facilitate effective learning experiences. However, in the learning process there is a problem for teachers to know the level of student attention in the classroom to the material presented, so that the teacher does not know accurately the concentration of students during the learning process until it has an impact on the teacher's learning methods that are not in accordance with the characteristics of students. The purpose of this research is to detect students' facial expressions in the classroom learning process using yolov7. The implementation of several architectural models on CNN consists of several proposed methods, namely data collection, data augmentation, data annotation, split dataset, training, and model evaluation. System testing is done by measuring accuracy and comparing with other methods, namely CNN, CNN MobileNet, CNN EfficientNet-B0 and YoloV7. The test results show the average accuracy of CNN 80%, CNN MobileNet 93%, CNN EfficientNet-B0 31% and YoloV7 96%. Based on these results, it can be concluded that the YoloV7 method can detect student concentration effectively and efficiently compared to CNN, CNN MobileNet, and CNN EfficientNet-B0.

Banana and Orange Classification Detection Using Convolutional Neural Network

2024-07-25T12:17:59+00:00

Fruits play a crucial role in human health, with an average consumption of 81.14 grams per capita per day in Indonesia, where bananas and oranges are the most consumed fruits. Inconsistent fruit quality, typically evaluated manually by farmers, can influence consumer decisions. Artificial intelligence (AI) and computer vision can enhance efficiency and consistency in analyzing fruit quality. Convolutional Neural Networks (CNN) are particularly effective in image recognition. This research uses CNN to classify the quality of bananas and oranges from a dataset of 4000 images, divided into 10% test data, 80% training data, and 10% validation data. Among three models tested, Model 2 performed best with an accuracy of 96.75% and balanced high F1-scores across all categories. The results demonstrate that the CNN model is capable of classifying the quality of bananas and oranges with high accuracy and good evaluation results.

Analysis of Field Work Practice Information System Service Quality Using The Webqual 4.0 Method and Importance Performance Analysis

2024-07-26T02:30:41+00:00

In the current digital era, the quality of website services is a crucial factor in supporting the effectiveness and efficiency of information systems, including the Information Systems Study Program Field Work Information System (SIPKL) at Universitas Terbuka. However, currently there is no in-depth evaluation of the quality of SIPKL services from a user perspective. This research aims to review the service quality of the SIPKL website as a whole and measure the level of user satisfaction with the services provided. To achieve this goal, the WebQual 4.0 method is used which measures three main dimensions of service quality, namely usability, information quality, and interaction quality. In addition, the Importance Performance Analysis (IPA) method is applied to evaluate the importance and performance of each service attribute being measured, so as to identify areas that require improvement. Data was collected through a survey with 100 respondents from Information Systems study program students who had used the SIPKL website. The research results show a value of 101.6% for the level of conformity, which indicates that the SIPKL website service performance has met or even exceeded user expectations and interests. Meanwhile, the gap value is categorized as “Good” with a positive value of 0.08 or >0. Indicators that require improvement are in quadrants II and III. Overall, this research provides strategic recommendations for SIPKL website managers to improve service quality so that it is more optimal in supporting students' needs in undergoing PKL.

Developing a Marker-Based AR Application to Introduce Temples and Cultural Heritage to Younger Generations

2024-06-21T12:35:25+00:00

Preserving Balinese cultural heritage is crucial for sustaining community identity. In Bali, temples (pura) are central to spiritual and cultural life. However, younger generations, especially temple caretakers of Pemerajan Agung Sakti Padangsambian, are increasingly losing knowledge of these sacred spaces, weakening their sense of belonging, to preserve cultural traditions. Current media efforts has failed to engage this demographic. This research addresses this challenge by developing an application-integrated images compiled into books and Android-based AR technology. The application employed a user-centered design approach involving analysis, design, development, testing, and evaluation phases. Results show AR effectively bridges the knowledge gap, with usability scores and a significant increase in user knowledge of 42.43%. This research demonstrates AR's potential for preserving and transmitting cultural heritage, including the reconstruction of damaged historical objects through 3D modeling with the marker detection technology, to ensure seamless integration between the real and virtual worlds.

Correlation Analysis Approach Between Features and Motor Movement Stimulus for Stroke Severity Classification of EEG Signal Based on Time Domain, Frequency Domain, and Signal Decomposition Domain

2024-10-09T00:06:43+00:00

The healing process of a stroke necessitates tools for measuring relevant parameters to facilitate monitoring, evaluation, and medical rehabilitation. Accurate parameter measures can be observed in stroke patients' severity to ascertain suitable interventions by identifying components pertinent to monitoring, evaluation, and medical rehabilitation. The components are derived from the observation collection process utilizing an EEG device, accompanied by a motor stimulus, to ensure the acquisition of EEG signals for monitoring, evaluation, and medical rehabilitation while preventing any loss of information during data collection. The acquired information encounters challenges due to the signal's unstable, nonlinear, and non-stationary characteristics, necessitating efforts to stabilize, render stationary, and linearize it through suitable signal processing and feature extraction techniques to achieve a pertinent feature composition. The subsequent difficulty is achieving the objectives of medical monitoring, evaluation, and rehabilitation, necessitating the correlation between EEG signal characteristics and motor movement stimuli, ensuring that the process adheres to appropriate parameter identification and scheduling per the established plan. In response to this difficulty, a correlation analysis methodology is established, incorporating normalcy tests, significance tests, and correlation analysis to ensure that the relevant factors for identifying stroke severity categorization patterns are precisely identified beforehand. The correlation analysis strategy employs raw data situations, preprocessing, feature extraction, feature selection, and correlation analysis for classification purposes. Our experimental findings indicate that the correlation analysis approach for assessing stroke severity classification patterns is evident in the Hajorth Complexity feature, utilizing the Shoulder motor movement stimulus and the SVM classification type, achieving an accuracy significant value of 98%. These findings confirm the efficacy of correlation analysis between EEG signal features and motor movement stimuli in identifying the optimal parameters within a reduced dimensional space to assess stroke severity effectively.

Enhancement of Internal Business Process Using Artificial Intelligence

2024-07-26T01:15:44+00:00

This research aims to explore the feasibility of Artificial Intelligence (AI) enabled process improvement systems to assist businesses in optimizing Internal Business Process (IBP) by making and adopting suggestions and improvements. Over the last two decades’ technological advances in the new generation have allowed us to use more sophisticated systems to speed up different tasks, as well as AI which has cognate from theory to something more efficient and applicable. This study confirms that a feasible AI-based system can provide benefits to companies in terms of increasing revenue. A mixed method was used; quantitative research was carried out through surveys to gather knowledge about the use of AI in the IBP, while qualitative research was carried out through interviews to obtain an overview of the use of AI in certain IBP. The results show that constructing AI in process optimization is a complicated task than one might expect.

Optimizing The User Interface of Waste Bank Application Using UCD and UEQ

2024-08-14T01:49:57+00:00

Environmental cleanliness is an essential aspect of life to make a healthy and comfortable environment. In Indonesia, the volume of waste will reach 70 million tons by 2022, with around 24% or 16 million tons needing to be appropriately managed. Related to the significant waste growth, the Ministry of Environment has developed the Waste Bank initiative, a collaborative effort that aims to educate the public in sorting waste and raising awareness of the importance of wise waste management. The desire of the local environmental agency to connect with the community supports the researcher in developing the Waste Bank application. The application will implement an optimal User Interface (UI) and User Experience (UX) design. The User-Centered Design (UCD) method will be employed, supported by the User Experience Questionnaire (UEQ), and is used to design UI and UX for the Waste Bank mobile application. The application prototypes were tested and evaluated using UEQ. The first design achieved an average score but still required improvement. In contrast, the second design scored excellently in six aspects measured: attractiveness, perspicuity, efficiency, dependability, stimulation, and novelty, with significant improvement. These results show that the UCD and UEQ methods are effective for developing UI/UX designs to meet user needs and can be applied in mobile application developments.

Model GHT-SVM for Traffic Sign Detection Using Support Vector Machine Algorithm Based On Gabor Filter and Top-Black Hat Transform

2024-07-12T00:49:41+00:00

A factor that can hinder the detection and recognition of traffic signs is the variation in lighting in the image of traffic signs. This study aims to detect traffic symbols using Gabor Filter (GFT), Top Hat Transform (THT), and Black Hat Transform (BHT) methods on the Support Vector Machine (SVM) algorithm for traffic sign dataset images with data problems that tend to have dark backgrounds at night and bright backgrounds during the day. From the experimental results, GHT-SVM gets the highest accuracy compared to HSV-SVM, HSV-RF, HSV-KNN, and H2T-SVM models. Based on experimental results, H2T-SVM from HOG ⊕ ENT ⊕ BHT ⊕ SVM results get the best accuracy of 86.42%. The Gabor Filter (GFT) parameters used are the number of filters with a value of 16, ksize with a value of 30, sigma with a standard deviation value of 3.0, lambd with a sinusoidal factor value of 10.0, gamma with a spatial aspect ratio value of 0.5 and psi with a phase offset value of 0 while the Top Hat Transform (THT) and Black Hat Transform (BHT) methods use filterSize sizes with values (3, 3).

Early Detection Depression Based On Action Unit and Eye Gaze Features Using a Multi-Input CNN-WoPL Framework

2024-10-15T02:44:58+00:00

Depression is a common mental disorder with significant life impact, including a high risk of suicide. Patients with depression attempt suicide five times more often than the general population. Self-reporting, subjective judgement and clinician expertise influence conventional diagnostic methods. For timely intervention and effective treatment, early and accurate diagnosis of depression is essential. This study proposes a framework called Multi-Input CNN-WoPL, a CNN-based method without a pooling layer that combines two features - action units and gaze - to improve accuracy and robustness in automatic depression detection. Pooling layer reduces spatial dimension of feature map, resulting in loss of information related to expression data, affecting depression detection result. The performance of the proposed method results in an accuracy of 0.994 and F1 score = 0.993, the F1 score value close to 1.0 indicates that the proposed method has good precision, recall and performance.

Semantic Approach for Digital Restoration of Balinese Lontar Manuscripts

2024-09-20T01:11:57+00:00

Balinese lontar manuscripts represent a cultural heritage containing significant historical, religious, and scientific values. However, their centuries-old age makes them vulnerable to damage. This research proposes a semantic-based digital restoration solution to address this issue. The semantic approach comprehends the meaning and structure of the lontar text, ensuring an accurate restoration process that preserves the original meaning. The development of the semantic-based digital restoration is built using the Design Science Research Methodology (DSRM). The system is equipped with data management features that accommodate new data, ensure accurate information updates, and maintain the integrity of relationships between entities. Testing through SPARQL query approaches and black-box testing indicates that data additions, deletions, and modifications function well without conflicts or inconsistencies. Moreover, the system performs as expected and is ready for use. The implications of this research suggest that semantic-based digital restoration can be an effective solution for preserving Balinese lontar manuscripts and similar cultural heritage.

Systematic Literature Review: Use of Augmented Reality as A Learning Media: Trends, Applications, Challenges, and Future Potential

2024-07-12T00:42:07+00:00

This article conducts a systematic literature review (SLR) focusing on the application of Augmented Reality (AR) as an educational tool. The review process, guided by SLR and PRISMA methodologies, included steps such as identification, screening, eligibility assessment, inclusion, and data analysis, utilizing tools like Publish or Perish 8 and NVIVO 12 Plus. An initial search on Scopus produced 800 articles, which were subsequently narrowed down to 59 relevant studies. These were analyzed with NVIVO 12 Plus according to specific topics. The results indicate that AR effectively enhances students' academic achievement, interest, motivation, and participation across various subjects such as science, mathematics, languages, and engineering education. However, challenges include hardware and software limitations and insufficient technical training for teachers. AR holds great potential for improving learning experiences, particularly for students with special needs. Future developments should focus on affordable software and adequate teacher training to expand AR's educational use. Further research should explore AR in vocational education to better understand its specific requirements.

Optimizing Diabetic Neuropathy Severity Classification Using Electromyography Signals Through Synthetic Oversampling Techniques

2024-11-02T02:51:14+00:00

Electromyography signals are electrical signals generated by muscle activity and are very useful for analyzing the health conditions of muscles and nerves. Data imbalance is a prevalent issue in EMG signal data, especially when addressing patients with varied health conditions and restricted data availability. A major difficulty for machine learning models is class imbalance in datasets, which frequently leads to biased predictions favoring the dominant class and neglecting the minority classes. The data augmentation method employs the Synthetic Minority Over Sampling Technique (SMOTE) and Random Over Sampling (ROS) to address data imbalances and enhance the performance of classification models for underrepresented classes. This study employs an oversampling technique to enhance the efficacy of the XG Boost model. SMOTE exhibits better efficacy relative to competing methods; the application of appropriate oversampling techniques allows models to integrate patterns from both majority and often neglected minority data.

Real Time Automated Speech Recognition Transcription and Sign Language Character Animation on Learning Media

2024-11-05T00:50:53+00:00

Inclusive education for deaf students requires a technology approach to address communication and comprehension challenges. This study aims to develop innovative learning media that integrates real-time ASR (Automated Speech Recognition) transcription technology and sign language character animation to improve accessibility and comprehension of materials for deaf students. This learning media receives input from live voice, voice from learning videos, and text inputted by teachers. Using the Google Cloud API-based ASR transcription module, voice and text are converted into written text, broken down into vocabulary for sign language animation search. The search is carried out using an interpolation algorithm in the sign language animation asset database, allowing the display of animations relevant to the transcribed vocabulary.

The development process follows the ADDIE instructional design model, starting with needs analysis and ending with implementation and evaluation. The analysis stage includes data collection through teacher interviews, classroom observations, and curriculum reviews. The media design is designed to meet the specific needs of deaf students, while development and implementation focus on technology integration and effective material delivery. Evaluation is carried out to assess the effectiveness of the media in improving student understanding and participation. The study's results showed that this learning media can improve deaf students' understanding of the material and increase their involvement in the learning process. ASR technology and sign language animation contribute significantly to making learning materials more accessible and understandable, supporting the goals of inclusive education.

Implementation of a Web-Based Master-Slave Architecture for Greenhouse Monitoring Systems in Grape Cultivation

2024-08-14T03:53:08+00:00

The Internet of Things (IoT) technology enables electronic devices to connect to the Internet for real-time data collection and analysis. In greenhouses, IoT is used to monitor soil moisture and environmental conditions to support grape plant care. This study proposes a grape plant monitoring system using a master-slave architecture and the ESP-NOW protocol to reduce reliance on Wi-Fi networks, thus minimizing delay and packet loss. The system leverages direct communication between master and slave nodes. Testing results show an average delay of 1,546.65 ms, jitter of 120.56 ms, and packet loss of only 0.07% from 88,815 data transmissions in one day. Despite variations in packet loss due to power interruptions, the system consistently demonstrates reliable data transmission. Overall, this system proves to be reliable for real-time monitoring in greenhouses, offering stable performance and high data accuracy.

Synthesis of Kantil Tone Using The Frequency Modulation Method

2024-09-11T07:10:37+00:00

Music is a creative form of expression that utilizes sound arranged in specific patterns to create artistic works that are enjoyable to the listener. However, in music, excessive or continuous exposure to loud sounds can damage the hair cells in the ear, potentially leading to hearing loss or even deafness. One challenge in musical instrument craftsmanship is the variation in sound produced by different kantil artisans. These differences in sound output lead to inconsistencies in the rhythm of the angklung gamelan in Bali. This research addresses the issue by focusing on the process of synthesizing kantil sounds to achieve a more consistent output. The research begins by inputting audio files for each sound bar in format. The recorded audio data undergoes preprocessing using the Fast Fourier Transform (FFT) method, which extracts key features from the dataset, such as the fundamental frequency. Additionally, the Hilbert Transform is applied to obtain the optimal sound each blade, which will later be used in the Frequency Modulation process. Once preprocessing is completed on the dataset for each blade, the fundamental frequency and signal are acquired. To evaluate the accuracy of the synthesis, the Root Mean Square Error (RMSE) is calculated to compare the original signal with the synthesized signal. This step helps determine the degree of difference between the two signals. Ultimately, the result is a synthesized kantil sound that closely resembles the original, helping to standardize sound output among different craftsmen and ensuring consistency in musical performances

Optimization of Sales Data Forecasting Computation Process Using Parallel Computing in Cloud Environment

2024-11-05T01:37:04+00:00

The Holt-Winters Exponential Smoothing algorithm optimised using the Modified Improved Particle Swarm Optimization (MIPSO) algorithm is an algorithm that is able to provide good sales data forecasting results. However, there is a problem that when the iteration process is carried out using 1 computer, it takes a long time to finally get the test results. It is necessary to optimise the computational process to get more optimal and efficient results. This research will combine parallel computing technology and cloud computing technology to help speed up the computing process. The results of this research show that the more server used, the greater the reduction in execution time that occurs, because heavy computing tasks can be distributed more efficiently to many machines. This is evident from the comparison between single server and parallel server. Then the combination of more cores and servers produces the most optimal configuration in accelerating computation.

Usability and Performance Comparison: Implementation of Tibero and Oracle Databases in the Context of CAMS Software Development

2024-08-15T04:33:03+00:00

In the world of software development, the role of database systems is very vital. Enterprise software, designed to handle the complex challenges of large organizations, requires reliable and efficient databases. Oracle, one of the top choices in the industry, stands out with its performance and flexibility. On the other hand, Tibero, a relational DBMS from TmaxSoft, offers the high performance, reliability and scalability required in business environments that require big data management. This research was conducted to analyze the technical side of the Oracle and Tibero databases in the context of the CAMS (Customer Asset Management System) application, with a focus on usability and performance aspects. This research uses the Performance Testing method to evaluate CPU, Memory, Storage resource usage and TPS (Transaction Per Second) of the two databases as well as the System Usability Scale (SUS) to measure user experience. The results provide information to software developers in selecting databases that suit business needs, while contributing to the development of the information technology industry

Smart Home for Supporting Elderly Based On Ultrawideband Positioning System

2024-09-18T02:56:56+00:00

In 2017, the rate of dependency among the elderly was reported to be at 13.28%, which was problematic, due to the limited number of caregivers to assist them at all times. To address this issue, a robotic service and vital sign-based system were developed, but it was found to be insufficient for monitoring the activities of the elderly. Therefore, this study aimed to address the high dependency rates of elderly individuals who required constant support and care to survive by designing an ultrawideband-based positioning system. The system consisted of five sub-systems, including an indoor positioning system, a database system, a data processing system, an actuator system, and an application user interface. The system testing phase revealed several important findings, including that the position coordinates of the elderly were accurately read with differences of only 98.884 mm and 279.94 under Line of Sight and Non-Line of Sight conditions, respectively. Furthermore, the initial error rate of 164.39% was successfully reduced to only 1.096% by applying the average filter method in the data processing system. The actuator system also showed an impressive accuracy rate of 98% success, while the Android-based application user interface received a high user experience rate of 92.3%. Overall, these findings suggested that the ultrawideband-based positioning system had significant potential to support smart homes for the elderly and improve their quality of life.

The Influence of Educational Robotics in STEM Integrated Learning and the Development of Computational Thinking Abilities

2024-07-06T21:14:29+00:00

Currently, educational robotics are becoming an important trend in education, introducing transformative elements into the classroom to improve the learning environment. Educational robotics in STEM-integrated learning can develop computational thinking skills. Educational robotics has begun to be widely adopted and is expected to enhance computational thinking skills in early childhood education, secondary school, and higher education. In this study, we examine the role of educational robotics in integrated STEM learning environments and its impact on the development of computational thinking. The method used was a systematic literature review. Initial search returned 541 articles from various journals indexed in Scopus. Subsequently, 351 articles published between 2020 2024 were sorted out, and only 37 articles were included in the final analysis. Studies show that educational robotics effectively promotes STEM education and facilitates the development of computational thinking skills. The importance of project-based learning and the integration of STEM disciplines in educational robotics inform educators and policymakers about the potential benefits of educational robotics in promoting STEM education and developing computational thinking skills.

Optimization of XGBoost Algorithm Using Parameter Tunning in Retail Sales Prediction

2024-08-12T14:24:44+00:00

In retail companies, the owner needs sales analysis to make decisions in the company's business processes. Several previous studies have introduced forecasting techniques using regression analysis, and classification approaches that need optimization. This article proposes a new approach to sales prediction using XGBoost, which is optimized by comparing the best performance from three optimization methods: Random search, grid search, and Bayesian optimization. The aim is to obtain the best comparative analysis and increase prediction accuracy. The novelty of the proposed model is determining the best value for each optimization method using XGBoost. The results of the evaluation show that the best results were achieved by the grid search optimization technique in the XGBoost model with an increase in the evaluation value R^2 from 97.31 to 98.41. The results of the proposed model analysis can help retail business owners in accurate sales predictions to determine the development of business processes.