Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control
Not a member yet
424 research outputs found
Sort by
Predicting the Sentiment of Review Aspects in the Peer Review Text using Machine Learning
Get PDFThis paper develops a Machine Learning (ML) model to classify the sentiment of review aspects in the peer review text. Reviewers use the review aspect as paper quality indicators such as motivation, originality, clarity, soundness, substance, replicability, meaningful comparison, and summary during the review process. The proposed model addresses the critique of the existing peer review process, including a high volume of submitted papers, limited reviewers, and reviewer bias. This paper uses citation functions, representing the author's motivation to cite previous research, as the main predictor. Specifically, the predictor comprises citing sentence features representing the scheme of citation functions, regular sentence features representing the scheme of citation functions for non-citation sentences, and reference-based representing the source of citation. This paper utilizes the paper dataset from the International Conference on Learning Representations (ICLR) 2017-2020, which includes sentiment values (positive or negative) for all review aspects. Our experiment on combining XGBoost, oversampling, and hyper-parameter optimization revealed that not all review aspects can be effectively estimated by the ML model. The highest results were achieved when predicting Replicability sentiment with 97.74% accuracy. It also demonstrated accuracies of 94.03% for Motivation and 93.93% for Meaningful Comparison. However, the model exhibited lower effectiveness on Originality and Substance (85.21% and 79.94%) and performed less effectively on Clarity and Soundness with accuracies of 61.22% and 61.11%, respectively. The combination predictor was the best for the 5 review aspects, while the other 2 aspects were effectively estimated by regular sentence and reference-based predictors
Spatial Interpolation Long-Term Patterns Capacity of Solar Energy in Sumatera
Get PDFIndonesia possesses considerable capacity for renewable energy as a result of its plentiful natural resources, including geothermal, solar, wind, hydro, and biomass. However, the nation's existing energy composition is predominantly dependent on non-renewable resources, with fossil fuels constituting approximately 95% of its overall energy consumption. Recently, Indonesia has made notable advancements in augmenting its renewable energy output in years. Nevertheless, there is still obscurity about the identification of suitable regions for the installation of solar power plants in order to facilitate the development of solar energy. This study employed a methodology to investigate and forecast the solar energy potential in Sumatra, Indonesia. The data utilized consists of MERRA-2 reanalyzing information spanning from 1980 to 2019, collected on a daily basis. The data is analyzed and shown using Inverse Distance Weighting and ARIMA techniques to visualize the spatial variation of solar energy potential in Sumatra. ARIMA is employed as a supplementary method to the interpolation technique in order to get long-term projections of solar energy potential for the period spanning from 2020 to 2029. The analysis of the best interpolation method for estimating solar energy potential reveals that the IDW approach with a power of 5 yields the most accurate findings, with an RMSE value of 28.33. For long-term prediction of solar potential in Aceh province, the ARIMA (1,0,0) method is recommended, which has a MAPE value of 0.0219. The findings indicated that Lampung and Bengkulu frequently experience the distribution of solar energy with an intensity ranging from 1400 to 1450 kWh. In addition, the forecast of the potential over Sumatera Island yielded encouraging results using the GAM model, with a root mean square error rate of 0.05103
Tomato Leaf Diseases Classification using Convolutional Neural Networks with Transfer Learning Resnet-50
Get PDFThis research delves into the critical domain of Tomato Leaf Disease classification using advanced machine learning techniques. Specifically, a comparative evaluation was conducted between a Base CNN model devoid of ResNet-50 integration and a Proposed Method harnessing the capabilities of ResNet-50. The results elucidated a notable enhancement in performance metrics when leveraging ResNet-50, with the Proposed Method consistently achieving exceptional accuracy scores of 99.96%, 99.98%, and 99.96% across data splits of 90:10, 80:20, and 70:30, respectively. Furthermore, the ResNet-50 integration significantly augmented key metrics, including recall, precision, and F1-Score, thereby accentuating its pivotal role in enhancing sensitivity and positive predictive value for tomato leaf disease classification. As for prospective research trajectories, this study highlights potential avenues for refinement, encompassing the exploration of ensemble techniques amalgamating diverse architectural frameworks, advanced data augmentation methodologies, and broader disease classification scopes. Collectively, this research underscores the transformative potential of ResNet-50 in agricultural diagnostics, advocating for continued exploration and innovation to fortify global food security and sustainable farming practices. Future research could explore ensemble techniques, advanced data augmentation, broader disease classification scopes, and interdisciplinary collaborations to develop comprehensive diagnostic tools for sustainable farming practices and global food security
Smart AODV Routing Protocol Strategies Based on Learning Automata to Improve V2V Communication Quality of Services in VANET
Get PDFThe Adhoc On-Demand Distance Vector (AODV) protocol faces challenges in selecting the best relay nodes, which requires optimization to improve performance in Vehicular ad-hoc networks (VANETs). This study aims to enhance Vehicle-to-Vehicle (V2V) communication in VANETs by implementing the Learning Automata-Driven Ad-hoc On-Demand Distance Vector (LA-AODV) routing protocol. LA-AODV is designed to achieve higher packet delivery ratios and optimize data transfer rates, even under congested network conditions, by dynamically adjusting to changing network scenarios. The performance evaluation includes six key metrics analyzed under varying node densities and time intervals, comparing LA-AODV against the standard AODV protocol. Results indicate that LA-AODV consistently outperforms AODV, demonstrating improved efficiency in flood identifier management, reduced data loss, higher packet delivery ratios, better throughput, and reduced end-to-end delay and jitter. Specifically, under a 20-node scenario, LA-AODV exhibits lower flood ID scores (54 vs. 88), reduced packet loss (11% vs. 12%), higher PDR (88.0% vs. 87.0%), and superior throughput (85.34 Kbps vs. 47.26 Kbps). Additionally, LA-AODV achieves lower end-to-end delay (6.84E+09 ns vs. 3.76E+10 ns) and jitter (2.52E+09 ns vs. 2.15E+10 ns). These findings suggest that LA-AODV significantly enhances Quality of Service (QoS) in vehicular ad-hoc networks, positioning it as a promising solution for optimizing V2V communication performance
Improving Software Defect Prediction Using a Combination of Ant Colony Optimization-based Feature Selection and Ensemble Technique
Get PDFSoftware defect prediction plays a vital role in enhancing software quality and minimizing maintenance costs. This study aims to improve software defect prediction by employing a combination of Ant Colony Optimization (ACO) for feature selection and ensemble techniques, particularly Gradient Boosting. This research utilized three NASA MDP datasets: MC1, KC1, and PC2, to evaluate the performance of four machine learning algorithms: Random Forest, Support Vector Machine (SVM), Decision Tree, and Naïve Bayes. The data preprocessing comprised handling class imbalance using SMOTE and converting categorical data into numerical representations. The results indicate that the integration of ACO and Gradient Boosting significantly enhances the accuracy of all four algorithms. Notably, the Random Forest algorithm achieved the highest accuracy of 99% on the MC1 dataset. The findings suggest that combining ACO-based feature selection with ensemble techniques can effectively boost the performance of software defect prediction models, offering a robust approach for early detection of potential software defects and contributing to improved software reliability and efficiency
BGP Dynamic Routing Protocol: A QoS Analysis for TCP and UDP
Get PDFThe Border Gateway Protocol (BGP) is commonly used for TCP and UDP services, but it poses challenges in terms of Quality of Service (QoS) analysis. Parameters like throughput, packet loss, delay, and jitter are crucial for assessing network service quality. This study aims to analyze the performance and influence of the BGP routing protocol on TCP and UDP services using QoS parameters. The research used a GNS3 network simulation to conduct multiple packet transmission tests for TCP and UDP protocols, lasting 15, 30, and 60 seconds; and monitored using Wireshark. For TCP services, the average QoS index value is 3.75, categorizing the quality as "Good". The tested network topology and routing configuration exhibit reliable performance, providing good throughput, low packet loss rates, minimal delays, and stable jitter. Similarly, UDP services demonstrate “Good” performance with an average QoS index of 3.75. The BGP routing protocol in the tested network topology ensures high-quality service with good delivery speed, low packet loss rate, minimal delay, and stable jitter. Overall, the study concludes that the BGP routing protocol effectively provides satisfactory QoS for TCP and UDP services. This research contributes to understanding network performance and optimizing routing protocols for improved telecommunications services. The findings highlight the significance of routing protocols in facilitating efficient data transmission on the Internet, reinforcing the importance of QoS analysis for enhancing service quality
Website Quality Analysis Using Modified Webqual Method and Importance Performance Analysis on SITU TAK Website
Get PDFTechnological developments affect information services, such as website. Information services on the website make it easy to convey information widely. Therefore, the quality of the website can affect the information services. This research assessed one of Telkom University's academic service websites, namely SITU Student Activities Transcript (SITU TAK). The purpose of this study was to measure the quality of the website, user satisfaction, and the factors that can increase the user satisfaction. This study employed Webqual 4.0 method as the indicator and Importance-Performance Analysis (IPA) for grouping the factors based on the quadrant of IPA. Before grouping the data, the data first passed the validity test, reliability test, and gap analysis between user perceptions and expectations. Therefore, it can strengthen the conclusions and recommendations resulted from this study. After conducting this study, the final results were obtained, which stated that SITU TAK website still did not meet the expectations of its users. This can be seen in the results of the gap analysis calculation with a value of -0.63, which means that the level of importance or expectations of the users is still higher than the performance of the website
Advancements in Cooperative Mobile Robots Control Strategies for Large-Scale Material Transport: Review
Get PDFThis paper explores groundbreaking advancements in control strategies for cooperative mobile robots used in large-scale material transport, a critical aspect of modern industrial, manufacturing, logistics, and construction sectors. It delves into the development of sophisticated systems that enable seamless coordination among multiple mobile robot systems. The research presents a novel hierarchical finite state automaton for dynamic mission adaptation and a null space-based control scheme for precise task execution and enhanced system resilience. The introduction of Mecanum wheels facilitates flexible movement and manipulation of materials, thereby increasing the operational efficiency and safety. Cutting-edge sensory technology, including LiDAR (Light Detection and Ranging), and the implementation of Robot Operating System are highlighted for their roles in enhancing autonomous navigation and intelligent operation. Additionally, the paper discusses the impact of centralized and decentralized control methods in ensuring safe cooperative object transport. The findings contribute to the vision of Industry 4.0 by promoting the integration of automation and robotic cooperation in complex environments and present a foundational blueprint for further research. Challenges for future work such as scalability, communication efficiency, collision avoidance, and energy efficiency are also considered, underscoring the need for ongoing development of robust and scalable robotic systems to address modern transport challenges
Layout Generation: Automated Components Placement for Advertising Poster using Transformer-based from Layout Graph
Get PDFIn the digital era, graphic design plays an important role in a company's marketing strategy, especially advertising posters that can convey messages to the audience. However, the process of creating attractive and informative posters takes a long time, especially the component placement on the layout. This research aims to develop a layout generator system that automatically places components on the layout using one of the transformer-based models. The transformer-based model used is a Graph Transformer with edge features called SGTransformer, which accepts input data as a graph. SGTransformer consists of several graph transformer layers that will calculate the attention of node and edge features on the input layout graph. A layout graph describes the spatial relationship between components in a layout. The SGTransformer model was trained by using advertising poster datasets collected from social media. The performance of the model were evaluated using the evaluation metrics commonly used in the layout generation domain such as Alignment, Overlap, Max IoU, and FID. The scores obtained from each evaluation metric are 0.025, 1.274, 0.325, and 8.575 respectively. The model evaluation results show that SGTransformer can produce structured and more diverse layouts although there are still challenges such as overlap between components. Code and other materials will be released at https://github.com/syahdeee/Layout-Generator
Thorax X-ray Image Segmentation Technique Using Four Variants of Thresholding Algorithm
Get PDFPneumonia is a respiratory infection caused by bacteria, viruses or fungi, and has been recognized as a fairly common and threatening disease. When diagnosing this disease, doctors usually also use thorax X-ray images. Nowadays, diagnosing pneumonia has been made possible with the help of machine learning technology. Doctors or medical personnel in locations where there are no pulmonary specialists or experts can be assisted by this technology. Machine learning algorithms are used to process digital images that have passed the pre-processing and segmentation stages. This paper offers a solution to segmentation technique of thorax X-ray digital image using a combination of four thresholding algorithms. This combination aims to find the best CNN model with segmentation techniques in the form of the most suitable thresholding algorithm. The result of this research is four different data sets. The thresholding algorithms used include binary, thresh binary inv, thresh to zero, thresh tozero inv with a threshold value of 150. The data used in this research is a thorax X-ray image dataset, as many as 5,856 images acquired from the Kaggle repository data. The program code in this research uses the Python programming language in the Anaconda environment. This research has resulted in a comparison of the accuracy values obtained using 4 variants between thres_binary thresholding algorithm and thres_binary_inv. The thres_tozero obtained 95% of accuracy while thres_tozero_inv obtained 94% of accuracy