International Journal on Advanced Science, Engineering and Information Technology
Not a member yet
2006 research outputs found
Sort by
XGBoost Classifier for DDOS Attack Detection in Software Defined Network Using sFlow Protocol
Get PDFFrom a security perspective, Software Defined Network (SDN) separates security concerns into Control Plane and Data Plane. The Control Plane is responsible for managing the entire network centrally. Centralized SDN generates high vulnerability against the Distributed Denial of Service (DDOS). When the Software Defined Network overwhelms by DDOS, both Control Plane and Data Plane will lack resources. It can cause the SDN to fail to work if not detected early. Using the ability of sFlow Protocol to capture the flow traffic in real time, the data could be used to detect DDOS attacks. This sFlow sampling approach can reduce the workload of the network by lower down the processing in switches. This paper uses Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM), and Random Forest as detection methods. We use ONOS as SDN Controller and build the topology in GNS3. Prometheus retrieves data from the sFlow Collector as a time series database. The classifier then uses the data from Prometheus for DDOS detection. For the dataset, we use four different datasets. Datasets 1 and 2 consist of 6109 data, each divided into two classes and three classes. Datasets 3 and 4 consist of 400488 data divided into 2 and 3 classes, respectively. The evaluation results have proved the effectiveness of the proposed method. XGBoost has the highest accuracy of another algorithm. The best accuracy is 99.84% using Dataset 4 as the training set
Robust Pose Estimation of Pedestrians with a Deep Neural Networks
Get PDFIn this paper, we provide a method for robust estimation of pedestrian pose that is especially useful for autonomous vehicles traveling toward pedestrians far away. Pedestrians in the far distance appear relatively small when seen by a camera, making it difficult to estimate the pedestrian's pose. We use fused deep neural networks (DNNs) to resolve the problems presented by pedestrians in the far distance. First, DNNs are used to detect pedestrians and enlarge the observed image. Next, the DNN method of pose estimation is applied. The proposed method uses a single camera to estimate the posture of a pedestrian in the far distance. Far-off pedestrians observed by cameras in moving cars appear as low-resolution images of non-rigid bodies. Detection and orientation estimation are difficult with conventional image processing methods. We used a series of DNNs to detect pedestrians, improve data availability, and estimate challenging postures to address these limitations. In this paper, we propose a method based on the multi-stage fusion of DNNs to solve a difficult problem for a single DNN. The experimental results established the superiority of the proposed method when applied to data challenging for conventional pose estimation methods. Applications of the proposed method include observing small objects and objects in the far distance. The method may be especially useful in surveillance systems, sports broadcasting, and other applications requiring human posture estimation
Adaptive Cone Algorithm
Get PDFThis study was conducted to promote a new adaptive cone algorithm (ACA) algorithm. ACA is a metaheuristic technique based on swarm intelligence. ACA contains three steps. Each agent moves closer to the global reference in the first step. Then, each agent searches for a better solution around the current solution in the second step. The global reference searches for better solutions around it in the third step. This algorithm is named cone because the local space size declines linearly during the iterative process. ACA introduces a new adaptability model to improve the exploration strategy when a better solution cannot be achieved. It is conducted by enlarging the local solution space. ACA is challenged to find the final solution for theoretical and practical problems. The 23 functions are chosen as theoretical optimization problems. The portfolio optimization problem is selected as the practical problem. ACA is compared with five algorithms: particle swarm optimization (PSO), grey wolf optimizer (GWO), marine predator optimization (MPA), average subtraction-based optimizer (ASBO), and pelican optimization algorithm (POA). The result shows that ACA is competitive in finding the optimal solution for 23 functions and outperforms all sparing algorithms in achieving the highest total capital gain in tackling the portfolio optimization problem. ACA is superior to PSO, GWO, MPA, ASBO, and POA in solving 20, 11, 13, 4, and 21 functions, respectively. In the future, ACA can be implemented in solving various practical optimization problems
Artificial Intelligence for the Classification of Plastic Waste Utilizing TinyML on Low-Cost Embedded Systems
Get PDFBCG's implementation of the economy makes Thailand more environmentally conscious. The consolidation policy encourages consumers to eliminate single-use plastics using the 3Rs. This article introduces a solution to reduce plastic waste drastically using artificial intelligence. Utilizing a low-cost Arducam Pico4ML embedded device and TinyML, a plastic waste classifying system prototype is developed for plastic bottle segregation. The grayscale image datasets of PET, HDPE plastic bottles, and unknown objects are adjusted in the image pre-processing state and utilized to create trained models using MobileNetV2 convolutional-based neural network algorithms. Effective feature extraction and model training are performed on the Edge Impulse platform, and the trained model is exported to an embedded device using the optimized compiler. A further RS485 Modbus communication protocol feature enables integration with a programmable logic controller (PLC). The validation results of the trained model indicate a classification performance of 100% accuracy. Based on the average precision results, it is notable that the trained model can recognize the most common waste with an average accuracy of over 90%. The minimum classification rate of the MobileNetV2 quantized model is 249 milliseconds. It is also implemented in low-cost embedded devices for real-time plastic waste classification using fewer processing resources (185.4K ROM and 88K RAM). The findings exhibit sequential contributions that satisfy the criteria for classifying plastic bottles and the machine's integration capacity. These outcomes are anticipated to foster social shifts in behavior and enhance public awareness about plastic waste management
Comparison of Crossflow Turbine Performance through Nozzle Position Variations Using ANSYS Simulation
Get PDFThe performance comparison of Crossflow turbines is greatly influenced by the position of the nozzle in the conversion of water energy into mechanical energy that occurs through the blades, runners, and shafts of Crossflow turbines. The study aims to directly examine the visualization of water fluid dynamics across the turbine runner blade and enhance the performance of the Crossflow turbine by varying the nozzle position. This study intends to investigate the impact of water flow dynamics and emission on the performance of Crossflow turbines with a combined horizontal-vertical nozzle position, specifically focusing on the magnitude of the number of turbine blades driven and the size of the runner blade area. The objective of investigating nozzle position variations in Crossflow turbines is to determine the specific nozzle position at which the turbine blade may efficiently extract maximum energy from the water flow, hence optimizing turbine performance. The research method using models made using CAD software is AutoCAD by exporting to IGES or IGS format to be compatible with ANSYS. The simulation of this research is with post-processing. There are three, namely making animations, making contours, and taking data to compare cross-turbine performance using variations in nozzle position. Crossflow turbine performance with horizontal nozzle position torque and turbine power is lower, and there is an increase in a vertical position. Then, the horizontal and vertical nozzle position is very good because the nozzle is more effective with maximum turbine performance, namely 13.811-watt turbine power 1,099 turbine torque at 120 rpm
Mental Health State Classification Using Facial Emotion Recognition and Detection
Get PDFAnalyzing and understanding emotion can help in various aspects, such as realizing one’s attitude, behavior, etc. By understanding one’s emotions, one's mental health state can be calculated, which can help in the medical field by classifying whether one is mentally stable or not. Facial Recognition is one of the many fields of computer vision that utilizes convolutional networks or Conv Nets to perform, train, and learn. Conv Nets and other machine learning algorithms have evolved to adapt better to larger datasets. One of the advancements in Conv Nets and machines is the introduction of various Conv architectures like VGGNet. Thus, this study will present a mental health state classification approach based on facial emotion recognition. The methodology comprises several interconnected components, including preprocessing, feature extraction using Principal Component Analysis (PCA) and VGGNet, and classification using Support Vector Machines (SVM) and Multilayer Perceptron (MLP). The FER2013 dataset tests multiple models’ performances, and the best model is employed in the mental health state classification. The best model, which combines Visual Geometry Group Network (VGGNet) feature extraction with SVM classification, achieved an accuracy of 66%, demonstrating the effectiveness of the proposed methodology. By leveraging facial emotion recognition and machine learning techniques, the study aims to develop an effective method
Benefits of Using Technology through the Use of Applications in Integrated Referral Services in Social Welfare Centers (Puskesos)
Get PDFIntegrated social services organized through the SLRT (Integrated Referral Service System) by the Puskesos (Social Welfare Center) were a step forward in answering today's increasingly complex social problems. On the other hand, the development of technology and information provided opportunities and challenges to unite the two technologies and integrate social services into a single unit to strengthen further efforts to solve social problems. The involvement of technology through applications was a step forward and quite visionary to take advantage of technological developments, especially the closer technology and information to today's society. Based on these conditions, the purpose of this study was to examine deeply the benefits of using technology in integrated referral services carried out by Puskesos. In line with these objectives, the method used was qualitative with descriptive type. Data collection techniques include documentation studies, observations, in-depth interviews, and focus group discussions. Furthermore, the sampling technique in this study was purposive sampling, and the number of informants in this study amounted to 70, spread over two locations, namely Sleman Regency and Bandung Regency. The results show four benefits that were quite dominant in the use of technology in integrated social services: effectiveness, efficiency, accountability, and public trust. Realizing this benefit, it is hoped that other supporting programs will be involved in the form of socialization related to the use of technology through the use of applications in these services
Analysis of 5.8 GHz Network for Line of Sight (LOS) and Non-Line of Sight (NLOS) in Suburban Environment
Get PDFThis paper presents the findings of radio wave characterization based on the measurement data at 5.8 GHz. The measurement data were collected by a testbed channel, which links with the following scenarios: a single tree, a row of trees, a row of trees and a road, a row of trees, a road, and a building. These experiments were conducted at University Teknologi Malaysia (UTM) Skudai, Johor to represent the suburban environment. The links consist of pairs of transmitting and receiving antennas that deploy the path of a line of sight (LOS) and non-line of sight (NLOS) radio propagation wave networks. Based on the measurement data analysis, the general issue concerning the statistical probability distribution and the characteristics of LOS and NLOS are examined and discussed. Note that 5.8 GHz technology can be used in both LOS and NLOS scenarios, but its performance varies based on the presence of obstacles and signal propagation characteristics. Other prominent experimental analysis methods, such as hypothesis testing and goodness of fit tests, are implemented to consolidate the findings. The analysis found that the empirical probability density function of LOS and NLOS channels follows Gaussian, Rayleigh, and Rician distribution. Predicting specific future technological developments, such as the availability of 5.8 GHz technology, is challenging because it depends on various factors, including research and development efforts, regulatory decisions, market demand, and technological advancements
Human Age Group Estimation Using Gait Features
Get PDFIn many practical applications, identifying the target age group is essential for marketing products and services. For instance, gaming and entertainment companies need to understand which age groups are most likely to purchase their services. This knowledge allows them to optimize their products and services to better cater to their target audience. This study proposes an age group prediction system using gait features. Gait, in this context, pertains to an individual's unique walking style. A diverse dataset containing subjects from 3 to 70 years old is collected. The age group is classified into three categories: child, adult, and senior. The critical aspect of this research lies in the preprocessing techniques applied to the gait patterns. The gait patterns are extracted from landmark human joint positions' key point values and preprocessed using smoothening techniques. Additionally, dimension reduction techniques enhance computational efficiency and accuracy before feeding the features into a deep learning-based classifier. These preprocessing steps play a pivotal role in the success of the deep learning-based classifier. A promising accuracy of up to 95% is reported for correctly recognizing the human age groups. The outcomes of this investigation underscore the tremendous potential of leveraging machine learning techniques to refine marketing strategies and boost customer satisfaction. The proposed approach can aid companies in aligning their products and services with the preferences and needs of distinct age groups, thereby enhancing their market presence and resonance with their target audience
The Formal Graph of APRDF
Get PDFA new alternative model for expressing more complex knowledge has been proposed as an attributed predicate RDF (APRDF). By handling attributes that represent any additional triples of the main triple, APRDF serves as a predicate. Therefore, the formal graph model of APRDF must be defined. Lastly, this work recommends that the APRDF's conventional diagram is a digraph-hypergraph mix. The previous formal graph of RDF is a hypergraph even though, visually intuitively, it is a digraph. It also contains inconsistency. The other new serialization needs to describe its formal model. Eventually, this work can provide the formal graph model of APRDF and maintain consistency. There have been a few definitions proposed. The direct impact of this formal model is that APRDF outperformed the other model significantly when retrieving complex queries within its formal graph. In querying, the initial implementation of the proposed formal graph takes an average of 62 milliseconds. Compared to the other graph models, the proposed formal graph can reduce query time by an average of 90,7 milliseconds on the BF-arch graph and 121,05 milliseconds on the naive/default graph. As the formal graph model is preserved, the attributed predicate of APRDF assumed will drive a new model in the retrieving process that more in using a predicate formed as a link in a graph. It will also be impacted in the mining process by more elaborate links/edges (link mining)