JOIV : International Journal on Informatics Visualization
Not a member yet
786 research outputs found
Sort by
Asana and Trello: A Comparative Assessment of Project Management Capabilities
Get PDFProject management tools are essential for streamlining project management activities and providing a variety of functionalities to assist organizations in executing projects efficiently. The selection of an appropriate tool is crucial, given the many options available in the market. This scholarly article employs a comparative analysis methodology to scrutinize two prominent project management tools, Asana and Trello. The aim is to assist companies and organizations in making informed decisions based on their specific needs. The comparative analysis delves into the strengths and weaknesses of Asana and Trello, assessing their features, functionalities, and suitability in the context of knowledge management areas. Both tools are evaluated for their capability to address project management challenges and improve organizational processes. The study concludes that the choice between Asana and Trello hinges on factors such as project scale, organizational requirements, and the preferred level of complexity. With its comprehensive features, Asana emerges as ideal for larger, agile-oriented projects. In contrast, Trello's simplicity and user-friendly interface suit relatively smaller projects well. This analysis provides valuable insights for organizations to align their project management tools with specific project conditions, facilitating optimising project execution processes to meet their unique goals and requirements. In terms of features, Asana outshines Trello by providing a more extensive range of functionalities that effectively support the mapping of knowledge management areas
Classification of Dermoscopic Images Using CNN-SVM
Get PDFTraditional machine learning methods like GLCM and ABCD rules have long been employed for image classification tasks. However, they come with inherent limitations, primarily the need for manual feature extraction. This manual feature extraction process is time-consuming and relies on expert domain knowledge, making it challenging for non-experts to use effectively. Deep learning methods, specifically Convolutional Neural Networks (CNN), have revolutionized image classification by automating the feature extraction. CNNs can learn hierarchical features directly from the raw pixel values, eliminating the need for manual feature engineering. Despite their powerful capabilities, CNNs have limitations, mainly when working with small image datasets. They may overfit the data or struggle to generalize effectively. In light of these considerations, this study adopts a hybrid approach that leverages the strengths of both deep learning and traditional machine learning. CNNs are automatic feature extractors, allowing the model to capture meaningful image patterns. These extracted features are then fed into a Support Vector Machine (SVM) classifier, known for its efficiency and effectiveness in handling small datasets. The results of this study are encouraging, with an accuracy of 0.94 and an AUC score of 0.94. Notably, these metrics outperform Abbas' previous research by a significant margin, underscoring the effectiveness of the hybrid CNN-SVM approach. This research reinforces that SVM classifiers are well-suited for tasks involving limited image data, yielding improved classification accuracy and highlighting the potential for broader applications in image analysis
Multi-Head Attention in Residual Networks to Improve Coral Reef Structure Classification
Get PDFResidual Networks (ResNet) mark a crucial advancement in convolutional neural network architecture, effectively tackling challenges like vanishing gradients for improved pattern detection in various image classification tasks. This study introduces a novel adaptation of the ResNet50 architecture that integrates a multi-head attention mechanism (MHA), coined MHA-ResNet50, for discerning coral reef structures within images. Strategic modifications are applied to the input of each stage, leading to the development of an MHA block, which is augmented by separable convolution. The deliberate inclusion of the MHA block at various stages in identity-block Resnet50, in adherence to multiscale gate principles, precedes its traversal through fully connected layers. Furthermore, we implemented the Stratified K-fold concept to ensure that each fold has a comparable proportion of each class. We successfully assessed the efficacy of the MHA-Resnet50 model in several MHA-block placement scenarios and saw improvements in the accuracy of coral reef structure predictions. The most optimal results were achieved by incorporating four attention blocks (MHA-ResNet50-4), yielding an accuracy rate of 85.23% in recognition of coral structure images, comprising a mere 409 images. This model showcases adaptability to small datasets while delivering commendable performance. The ResNet50 architecture undergoes enhancement in our proposed model by integrating multi-head attention, separable convolution, and multiscale gate principles. The MHA-ResNet50 model substantially advances accurately predicting coral reef structures, demonstrating adaptability to limited datasets. Future lines of this research involve digging deeper into the model design and using more significant amounts and classes of data to strengthen a more comprehensive range of generalizations
A Comparative Study of Feature Selection Technique for Predicting the Professional Tennis Matches Outcome in a Grand Slam Tournament
Get PDFTennis is one of the world's most played sports, attracting many spectators to participate in the game. One of the most essential strokes in a tennis match is serve performance. This research is intended to determine the most critical strokes in tennis serve performance in predicting the tennis match outcome. This research focuses on the Grand Slam Tournaments of the Australian Open, French Open, Wimbledon, and US Open. The data are collected on the tennis serve performances such as Percentage First Serve In (PFSI), Percentage First Serve Won (PFSW), Percentage First Serve Return Won (PFSRW), Aces, and many more. For one tournament, it consists of 254 observations. This study applied feature selection methods available in R programming, such as Correlation Matrix, Relative Importance Metrics, Boruta, MARS, and cForest. Selecting the most essential and correlated variables with the match status can improve the model and help produce better results. This might help the practitioners to apply this method to obtain the closest result to the actual outcome when we include the most correlated variables in the model. From the result obtained, variables of first and second serve, either win on serve or return serve, are identified as the most critical attributes in the tennis match. As a future implication, we suggest that these are all the factors the players need to pay extra attention to in winning the tennis match.
Coordination of The Apprenticeship Industrial Program with The Siakama Application
Get PDFThis research aims to examine the implementation of the SIAKAMA application in the Apprenticeship Industrial Program. This program was created as a SIAKAMA application to overcome hurdles during the monitoring and evaluation stages. At the monitoring stage, supervising lecturers and field supervisors can use the SIAKAMA application to monitor all Apprenticeship Industrial program student activities in the field, resulting in a good and smooth communication and coordination system. At the evaluation stage, the supervising lecturer and field supervisors in the SIAKAMA application can conduct assessments based on student activities in the field, including daily evaluations and final assessments after the Apprenticeship Industrial Program has been finished. This study employs a quantitative descriptive technique, the Research & Development method, and the 4D development model. A sample of Apprenticeship Industrial Program students from five departments of the Faculty of Engineering, Padang State University, was used in this study. The SIAKAMA application was found to be valid with a value of 0.876, practical with a value of 78.67, and effective with a value of 81.22% after data analysis using SPSS 25. This suggests that implementing the SIAKAMA application to enhance the work competency of Apprenticeship Industrial Program students is viable. The Apprenticeship Industrial Program model represents a modification of the Three Set of Actor development model, yet it hasn't been incorporated with the Industrial Revolution 4.0. Engaging in this Program enables students to acquire 4C skills, including Creativity and Innovation, Critical Thinking and Problem Solving, Communication, and Collaboration
Batik Classification using Microstructure Co-occurrence Histogram
Get PDFBatik Nitik is a distinctive form of batik originating from the culturally rich region of Yogyakarta, Indonesia. What sets it apart from other batik styles is its remarkable motif similarity, a characteristic that often poses a considerable challenge when attempting to distinguish one design from another. To address this challenge, extensive research has been conducted with the primary objective of classifying Batik Nitik, and this research leverages an innovative approach combining the microstructure histogram and gray level co-occurrence matrix (GLCM) techniques, collectively referred to as the Microstructure Co-occurrence Histogram (MCH).The MCH method offers a multi-faceted approach to feature extraction, simultaneously capturing color, texture, and shape attributes, thereby generating a set of local features that faithfully represent the intricate details found in Batik Nitik imagery. In parallel, the GLCM method excels at extracting robust texture features by employing statistical measures to portray the subtle nuances within these batik patterns. Nevertheless, the mere fusion of microstructure and GLCM features doesn't inherently guarantee superior classification performance. This research paper has meticulously examined many feature fusion scenarios between microstructure and GLCM to pinpoint the optimal configuration that would yield the most accurate results. The dataset used consists of 960 Batik Nitik samples, comprising 60 categories. The classifiers employed in this study are K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Decision Tree (DT), Naïve Bayes (NB), and Linear Discriminant Analysis (LDA). Based on the experimental results, the fusion of microstructure and GLCM features with the (LDA) classifier yields the best performance compared to other scenarios and classifiers
Preliminary Development of Vircadia Virtual Reality Platform for Monitoring Water Quality Powered by Solar Panels
Get PDFClimate change is a global issue that significantly challenges water resources, especially in regions with limited public awareness about water conservation. It manifests through rising global temperatures, shifting weather patterns, more frequent and intense natural disasters, and instability in water availability. These problems are worsened by low public awareness and the reliance on steam power plants for water pumps. Addressing these challenges requires educational media that raises awareness about the causes and impacts of climate change. This study introduces the early development of a Virtual Reality (VR) platform utilizing Vircadia, focused on creating a 3D world and monitoring water quality with the support of solar power. Vircadia, an open-source platform, offers developers the flexibility to build and host virtual worlds on their servers, providing greater control over the environment, scalability, and customization. With Vircadia, we can rapidly implement a VR platform that integrates custom assets from Blender and personalized avatars from Ready Player Me. Vircadia can seamlessly connect to IoT platforms via weblink, allowing for real-time monitoring of water quality parameters and enabling users to interact directly with and oversee IoT devices within the VR environment. This paper discusses why we chose Ready Player Me and Blender as platforms for building 3D avatars and assets, and Vircadia as the VR Platform. Additionally, it addresses challenges encountered when using Vircadia, such as asset optimization and IoT device integration. Future research will focus on optimizing asset quality, enhancing IoT integration, and implementing carbon emissions monitoring within the VR platform
Performance Analysis Of Machine Learning Algorithms Using The Ensemble Method On Predicting The Impact Of Inflation On Indonesia's Economic Growth
Get PDFThe warning of a global recession expected in 2023 is currently the world's concern. Global financial institutions have raised interest rates to lower inflation, which has led to this problem. This study aims to evaluate the effect of interest rates and inflation on Indonesia's economic growth and compare the performance of machine learning models, specifically Random Forest and XGBoost, in analyzing the impact of inflation. A qualitative methodology was used for the literature survey, while the quantitative approach involved the implementation of machine learning algorithms using the Ensemble Method. The results show that Random Forest performs better than XGBoost in predicting the impact of inflation on economic growth, with MSE values of 0.799 and 0.864 and MAE of 0.576 and 0.619, respectively. In addition, the R-squared value of Random Forest 0.908 is also higher than that of XGBoost 0.901, indicating that the model can better explain the variation in the target data. The practical implication of this study is that the Random Forest model can be more effectively used in analyzing the impact of inflation on Indonesia's economic growth. Recommendations for future research include exploring other methods and using more extended time series to deepen the understanding of the relationship between interest rates, inflation, and economic growth
Modeling and Application of Credit Scoring Based on A Multi-Objective Approach to Debtor Data in PT. Bank Riau Kepri
Get PDFThe development of information technology in Indonesia, marked by the start of Industry 4.0, is very rapid. With the development of technology, many companies use technology to develop their business, one of which is banking, which analyses the process of prospective customers. New employees find it challenging to interpret and tend to agree more easily with prospective customers because they only see the fulfillment of general requirements. This research aims to find an overview of the primary and additional factors to analyze prospective credit customers using The Cross-Industry Standard Process for Data Mining (CRISP-DM). Develop a model in this study using data variables of prospective customers in health insurance as a moderating variable. This model tested the Decision Tree algorithm with an accuracy value of 92.49%, the Random Forest with an accuracy value of 81.72%, the Support Vector Machine (SVM) with an accuracy value of 91.25%, and K-Nearest Neighbor (K-NN) with an accuracy value. 90.58%, Gradient Boosting with an accuracy value of 90.69%, and XGBoost with an accuracy value of 93.27%. The algorithm uses a cross-validation technique at the validation stage by changing the K value to 2, 4, 6, 8, and 10. The results show that the XGBoost Algorithm accuracy is 93.27% with a K value of 8. As the highest model accuracy, this model was implemented using the XGBoost Algorithm
Classification of Human Concentration Levels Based on Electroencephalography Signals
Get PDFConcentration denotes the capability to direct one's attention to a specific subject matter. Presently, within the era characterized by an overwhelming abundance of information inundating human existence, distractions frequently impede human concentration, thereby influencing the depth of knowledge acquisition. Various elements contribute to the decline in human concentration, including diminished metabolic states, inadequate sleep, and engaging in multiple tasks simultaneously. The cognitive state of an individual during the process of thinking can be assessed through the analysis of electroencephalography signals. The primary objective of this investigation is to facilitate experts' interpretation of electroencephalography signal outcomes for categorizing concentration levels. The dataset utilized in this examination comprises unprocessed EEG data obtained from observing individuals in both relaxation and concentration states. After data preprocessing, feature extraction is executed, and classification is performed using the Support Vector Machine technique. The outcome of this study reveals an accuracy rate of 84%. These developments allow for continual monitoring of brain function, an enhanced comprehension of cerebral activities, and increased operational efficacy of end-effectors. The implications of these advancements on prospective research opportunities are evident in the potential for more accurate diagnosis of neurological disorders and the progression of sophisticated BCI applications designed to support healthcare and monitor cognitive states. The evolution of EEG technology is paving the way for novel research pathways in neuroscience and human-computer interaction