Journal of Informatics And Telecommunication Engineering
Not a member yet
373 research outputs found
Sort by
Machine Learning-Driven Detection of Malicious URL: Comparative Analysis of Random Forest and SVMs
Get PDFNo longer a novelty, the internet has become the ubiquitous fabric of our lives, transforming how we interact, do business and disseminate information. However, its popularity has also attracted attackers who want to exploit it for personal gain. One tactic they use is to launch client-side attacks through malicious websites. Malicious websites are constantly evolving, and traditional methods such as blacklisting are no longer effective in identifying them. More sophisticated and adaptive solutions are needed to combat this threat. This research proposes an automatic malicious website detection method that utilizes URL properties and machine learning algorithms. This approach uses a combination of relevant URL features and a powerful machine learning model to accurately identify malicious websites. This research uses two popular machine learning algorithms: Random Forest (RF) and Support Vector Machines (SVM). Both models are trained on a dataset consisting of URL properties of malicious and Benign websites. The research results show that the proposed method is able to achieve a good level of accuracy in detecting malicious websites. Both RF and SVM show promising performance, with RF model achieved an accuracy of 86.15%, surpassing the SVM's performance of 85.38%. While overall performance is satisfactory, further optimization might be necessary, particularly to address potential class imbalance. Oversampling method could offer a more effective alternative to traditional undersampling methods and potentially improve performance across both website URLs categorie
Combination of Image Improvement on Segmentation Using a Convolutional Neural Network in Efforts to Detect Liver Disease
Get PDFLiver disease is a disease caused by various factors such as the spread of viruses. Liver damage causes the ability to break down red blood cells to be disrupted. Detection of liver disease can be done using the segmentation. Segmentation is useful for separating an area of the liver in an image from other areas. Segmentation carried out manually requires experts and a long time, so automatic segmentation is needed. CNN can be used to perform automatic segmentation. One of the CNN architectures is the U-Net architecture. Segmentation requires quality images to improve recognition of image patterns, so image improvement is needed in the form of contrast enhancement. Contrast improvement was carried out by taking Green Channel images. Contrast enhancement was carried out using the Contrast Stretching and CLAHE methods. The image improvement results show MSE and SSIM values 66.1844 and 0.7088. Evaluation of the image improvements obtained provides significant changes. The improved image is used at the segmentation stage. Segmentation is carried out using the U-Net architecture. The segmentation results obtained performance evaluation values in the form of accuracy 99.6%, sensitivity 98.9%, and specificity 99.7%. This shows that the proposed method can detect liver disease in liver images wel
Combination of Image Enhancement and U-Net Architecture for Cervical Cell Semantic Segmentation
Get PDFCervical cancer is the second leading cause of death in women and ranks fourth as a disease that occurs in women worldwide. Cervical cancer is a disease that is difficult to detect and can be detected when it is in an advanced stage. This requires early prevention by carrying out a pap-smear examination. Pap-smear examination manually requires a relatively long time, so a tool is needed by segmentation. Segmentation is image processing by performing perfection between the intended object and the background. One of the CNN methods commonly used in medical image segmentation is the U-Net architecture. Segmentation in this study was carried out on the nucleus and cytoplasm of the Herlev dataset using the U-Net architecture combined with data augmentation and image enhancement. In the learning process, this research resulted in a fairly high IoU value of 78% and an RMSE close to 20%. The results of this study also yielded an accuracy value of 89%, with an average precision, recall and F1 score of 89%, 89% and 88.67%, respectively. This shows that the combination of the CNN U-Net architecture with image quality improvement and data augmentation is quite good at segmenting cervical cells for the nucleus and cytoplas
Electronic Education Payment System Using Design Thinking Method At Mis Muhajirin
Get PDFElectronic Education Payment information system is an education fee payment system using online payment technology. However, the tuition payment system at MIS Muhajirin is still done manually by recording each payment transaction using the SPP card. This is an obstacle because it causes long queues by parents when paying tuition fees and there is also a high possibility of human error. This research aims to produce a web-based online payment information system at MIS Muhajirin, as well as to determine the level of efficiency if implemented at MIS Muhajirin. The method used in this research is the Design Thinking method which consists of 5 stages, namely Empathize, Define, Ideate, Prototype, and Test. The results of the research show that the information system developed provides effectiveness for admin staff when managing tuition payments at MIS Muhajirin and provides convenience for parents in terms of paying tuition online and not having to queue at school as before. Thus it can be concluded that this application is quite influential in improving the effectiveness and efficiency of the tuition payment system at MIS Muhajiri
Prediction Stunting Analysis Using Random Forest Algorithm and Random Search Optimization
Get PDFStunting is a nutritional problem experienced by toddlers, characterized by a height below the average. This condition arises due to various factors, one of which is the nutritional issues faced by toddlers. Stunting cases in Indonesia are relatively high, reaching 21.6% in 2022, indicating a significant prevalence of stunting. The identification of stunting is carried out through a data mining approach, deemed more efficient. However, the classification algorithm in data mining often encounters data imbalance, leading to low accuracy and inaccurate prediction results. To address this, the study employs the Random Forest algorithm with optimization using the random search method. The test results demonstrate that Random Forest achieves a relatively high accuracy of 90.7%. After optimization using random search, accuracy further increases to 96.33%. The combination of the algorithm and optimization proves to be highly effective, resulting in a 5.63% increase in accuracy. These findings hold crucial implications in supporting decisions for preventing stunting in toddlers. This research serves as a valuable source of information for the Health sector in identifying and implementing more effective strategies for stunting prevention. The use of the Random Forest algorithm optimized with random search proves to be an efficient solution in addressing data imbalanc
Analysis of the Level of Satisfaction with the Quality of Clinical Services Using the Servqual Method
Get PDFThe influence of the globalization era has increased the use of health facilities in society. This increase requires agencies to provide the best possible service to customers. UPT. The Sriwijaya University Health Clinic as one of the providers of health facilities has also experienced the impact of this increase, therefore this agency continues to provide the best service and makes improvements in the services provided so that service users can feel satisfied. The aim of this research is to determine the level of student satisfaction with services at UPT. Sriwijaya University Health Clinic and which dimensions of service quality have a dominant influence on student satisfaction levels. This research involved 117 respondents from all students and was distributed via Google form with a quantitative research type and survey approach. Data analysis techniques were carried out using SPSS 26 and calculations were carried out using the servqual method. The research results show that from the total, the level of student satisfaction is at a percentage of 87% in the very satisfied category and the quality dimension that has a dominant influence is the reliability dimension with a Gap value of -1.17. With Cartesian diagram analysis, the question instruments included in quadrant I and quadrant II have the opportunity to be improved and maintained in quality so that they can get better responses in the future. With this research, it can be taken into consideration by UPT. Sriwijaya University Health Clinic to improve the quality of services provided in every dimension, especially those included in the important category and maintain the quality of services that previously received a good respons
Automated Detection of Ancient Indonesian Coins for Historical Numismatic Investigations
Get PDFThis study focuses on cultural heritage preservation, employing the YOLOv8 model for detecting and classifying ancient Indonesian coins. Motivated by challenges in accurately identifying coins with diverse physical conditions, YOLOv8, introduced in January 2023, presents a groundbreaking advancement in computer vision. Meticulous testing on a curated dataset demonstrates the model's capacity to recognize unique features across various coin variations, marking a substantial stride in cultural heritage preservation. Contributing significantly to computer vision and cultural heritage studies, the research emphasizes key metrics in model evaluation—accuracy, recall, and precision. The rigorous assessment underscores the YOLOv8 system's efficiency and reliability in classifying ancient Indonesian coins, with experimentation yielding a commendable 91% accuracy. Beyond technical contributions, the study extends understanding of YOLOv8's capabilities, paving new avenues for object recognition in cultural preservation. This research serves as a milestone, advancing the integration of cutting-edge technology to safeguard and comprehend the rich cultural heritage embodied in ancient Indonesian coins
Modeling Of Hyperparameter Tuned RNN-LSTM and Deep Learning For Garlic Price Forecasting In Indonesia
Get PDFIn the Indonesian garlic industry, the unpredictability of garlic prices poses a substantial challenge, impacting the sector's stability and growth. This research aims to address this issue by developing a highly accurate predictive model using a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM). The study employs a dataset spanning 782 days, meticulously divided with 80% dedicated to training and 20% to testing. The model, equipped with 50 LSTM units, undergoes intensive training over 100 epochs, with a batch size of 5. Its effectiveness is evaluated using the Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE), revealing exceptional predictive capabilities. The model achieves a low RMSE and MAPE in both training and testing phases, underscoring its accuracy and reliability in forecasting garlic prices. These results indicate not only the success of the RNN-LSTM model in capturing the complex patterns of price fluctuations but also highlight the potential of machine learning in enhancing time series analysis. This breakthrough offers significant implications for stakeholders in the garlic industry, providing a powerful tool for informed decision-making and strategic market planning, thereby contributing to the sector's sustainable development and stabilit
Employee Attendance System with Facial Recognition Technology Using a Single Shot Detector (SSD) Algorithm
Get PDFThe attendance system is an important thing that is used to assess employee performance. Currently, many companies or organizations have developed attendance systems, but there are still many organizations that have not optimized attendance with a reliable system. This happens at the Palu City Tourism Office, where attendance still uses manual recording in album books. Many problems occur with the manual attendance process, so it is necessary to create an attendance system. This research aims to help the Palu City Tourism Office for attendance management by using face recognition technology using the Single Shot Detector (SSD) algorithm. The system development model uses Waterfall. This research uses 20 different facial poses with four people each with 5 poses. This research uses javascript programming language, Next JS framework, tailwind CSS, PostgreSQL database, and face-api.js library. Tests carried out with blackbox results all features on the system function properly then based on the results of face detection testing, the test results obtained that the Single Shot Detector algorithm can detect facial images recorded on the camera quickly and accurately in medium and bright lighting with a distance of 40-80 cm. The result of this research is a system that can recognize faces well by matching the training data image with the face image recorded on the camera in realtim
Performance Analysis Of The Combination Of Blum Blum Shub And Rc5 Algorithm In Message Security
Get PDFThis research aims to enhance message security in the RC5 algorithm by integrating it with the Blum Blum Shub (BBS) algorithm. The rapid growth in data and information exchange, driven by advancements in information and communication technology, demands robust security against attacks such as eavesdropping, interruption, and data modification. Cryptography, particularly with symmetric and asymmetric keys, becomes a solution to maintain message confidentiality. The RC (Rivest Cipher) algorithm, specifically RC5, has become a popular choice in network applications due to its speed and variable key length complexity. This study attempts to improve the quality of encryption keys by integrating the Blum Blum Shub (BBS) method, a mathematical random number generator algorithm. RC5 and BBS are used together to secure messages, producing ciphertext that is difficult to predict and smaller in file size compared to the standard RC5 method. The test results show that the processing speed is independent of the number of characters in the plaintext, while the encrypted file size resulting from the RC5-BBS combination is more efficient than using the default RC5. In conclusion, integrating BBS into RC5 can enhance the security and efficiency of the encryption algorithm, with the potential for widespread application in cryptography-based data securit