In this paper, we develop a novel feature selection method called RCFS (Regularized Copula based Feature Selection) based on regularized copula. We use l1 regularization, as it penalizes the redundant co-efficient of features and makes them zero, resulting in non-redundant effective features set. Scale-invariant property of copula ensures good performance in noisy data, thereby improving the stability of the method. Three different forms of copula viz., Gaussian copula, Empirical copula, and Archimedean copula are used with l1 regularization. Results prove a significant improvement in the accuracy of the prediction model than any non regularized feature selection method. The number of optimal features to achieve a fixed accuracy value is also less than any other non regularized feature selection techniques.
本文提出了一种基于正则化Copula的特征选择方法RCFS (regulalized Copula based feature selection)。我们使用l1正则化，因为它会惩罚特征的冗余系数并使它们为零，从而产生非冗余的有效特征集。copula的尺度不变性保证了该方法在噪声数据中的良好性能，从而提高了方法的稳定性。结合l1正则化，使用了三种不同形式的copula，即高斯copula，经验copula和阿基米德copula。结果表明，与任何非正则化特征选择方法相比，该模型的预测精度有显著提高。实现固定精度值的最优特征的数量也少于任何其他非正则化特征选择技术。
{"title":"An l1-Norm Regularized Copula Based Feature Selection","authors":"Snehalika Lall, S. Bandyopadhyay","doi":"10.1145/3386164.3386177","DOIUrl":"https://doi.org/10.1145/3386164.3386177","url":null,"abstract":"In this paper, we develop a novel feature selection method called RCFS (Regularized Copula based Feature Selection) based on regularized copula. We use l1 regularization, as it penalizes the redundant co-efficient of features and makes them zero, resulting in non-redundant effective features set. Scale-invariant property of copula ensures good performance in noisy data, thereby improving the stability of the method. Three different forms of copula viz., Gaussian copula, Empirical copula, and Archimedean copula are used with l1 regularization. Results prove a significant improvement in the accuracy of the prediction model than any non regularized feature selection method. The number of optimal features to achieve a fixed accuracy value is also less than any other non regularized feature selection techniques.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"125 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116691519","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In large-scale manufacturing operations with increasing global competition, success depends on a reliable network system to complete a value chain transformation. The use of the Industry 4.0 paradigm is increasingly evolving in many areas of different industries to ensure significant increases in factory productivity, flexibility, and efficiency. Consequently, the system transformation requires a shift from single automated node to a fully integrated system. However, selected methodology and results require study to fully understand the digital transformation as well as its characteristics. This investigation presents a system conversion study, between centralized and decentralized systems, using the concept of epidemic membership protocols in the context of Industry 4.0. This paper proposes the method based on membership protocols focusing on the system conversion methodology under dynamic network condition. The experimental results show that the proposed method provides an ability to rewrite the structure of the network topology with optimal accuracy of epidemic application.
{"title":"Epidemic System Conversion on Industry 4.0's Perspective Under Dynamic Network Condition","authors":"P. Poonpakdee, J. Koiwanit, C. Yuangyai","doi":"10.1145/3386164.3386180","DOIUrl":"https://doi.org/10.1145/3386164.3386180","url":null,"abstract":"In large-scale manufacturing operations with increasing global competition, success depends on a reliable network system to complete a value chain transformation. The use of the Industry 4.0 paradigm is increasingly evolving in many areas of different industries to ensure significant increases in factory productivity, flexibility, and efficiency. Consequently, the system transformation requires a shift from single automated node to a fully integrated system. However, selected methodology and results require study to fully understand the digital transformation as well as its characteristics. This investigation presents a system conversion study, between centralized and decentralized systems, using the concept of epidemic membership protocols in the context of Industry 4.0. This paper proposes the method based on membership protocols focusing on the system conversion methodology under dynamic network condition. The experimental results show that the proposed method provides an ability to rewrite the structure of the network topology with optimal accuracy of epidemic application.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124830088","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A system of anonymous processes that have no names assigned to them is considered in both synchronous and asynchronous communication models. The processes are fault free and can only communicate using test-and-set (TAS) registers. The aim of the paper is to assign unique names to all processes using a distributed algorithm. The naming of anonymous processes is studied in eight new problem models based on two categories; the number of TAS registers available, and the knowledge of the number of processes. In this paper, two distributed naming algorithms are developed that can assign unique names to anonymous processes. One is deterministic and the other is randomized. The developed algorithms are optimal in time complexity and namespace size. The Sequential Lookup algorithm, which is a deterministic algorithm, has a time complexity of 0(n2) steps, whereas the Random Lookup algorithm, which is a randomized algorithm, has a time complexity of 0(n log n) steps. Proof of the correctness of each naming algorithm is presented for all categories of the problem model where the number of processes is known. The Random Lookup algorithm has a better time complexity compared to the Sequential Lookup algorithm due to the use of randomness in accessing TAS registers.
{"title":"Naming Anonymous Processes with Test-and-Set Registers","authors":"Layla S. Aldawsari","doi":"10.1145/3386164.3386182","DOIUrl":"https://doi.org/10.1145/3386164.3386182","url":null,"abstract":"A system of anonymous processes that have no names assigned to them is considered in both synchronous and asynchronous communication models. The processes are fault free and can only communicate using test-and-set (TAS) registers. The aim of the paper is to assign unique names to all processes using a distributed algorithm. The naming of anonymous processes is studied in eight new problem models based on two categories; the number of TAS registers available, and the knowledge of the number of processes. In this paper, two distributed naming algorithms are developed that can assign unique names to anonymous processes. One is deterministic and the other is randomized. The developed algorithms are optimal in time complexity and namespace size. The Sequential Lookup algorithm, which is a deterministic algorithm, has a time complexity of 0(n2) steps, whereas the Random Lookup algorithm, which is a randomized algorithm, has a time complexity of 0(n log n) steps. Proof of the correctness of each naming algorithm is presented for all categories of the problem model where the number of processes is known. The Random Lookup algorithm has a better time complexity compared to the Sequential Lookup algorithm due to the use of randomness in accessing TAS registers.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"166 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126738499","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Recently there has been significant progress in the real-time monitoring system based on the Internet of Things (IoT). The use rate of IoT has been increasing exponentially because of its enormous application in different areas, with many of them are yet to be explored. This paper explains how to design an IoT system and describes its working mechanism. We present a general architecture of the real-time monitoring system using IoT and related services. We successfully implement our proposed architecture for a single domain. Then, we describe how to use the proposed architecture to monitor the different real-time contextual domains. Also, we present ideas on how to plug the data from a third-party application into the proposed architecture.
{"title":"A General Architecture for a Real-Time Monitoring System Based on the Internet of Things","authors":"Nilakantha Paudel, R. Neupane","doi":"10.1145/3386164.3387295","DOIUrl":"https://doi.org/10.1145/3386164.3387295","url":null,"abstract":"Recently there has been significant progress in the real-time monitoring system based on the Internet of Things (IoT). The use rate of IoT has been increasing exponentially because of its enormous application in different areas, with many of them are yet to be explored. This paper explains how to design an IoT system and describes its working mechanism. We present a general architecture of the real-time monitoring system using IoT and related services. We successfully implement our proposed architecture for a single domain. Then, we describe how to use the proposed architecture to monitor the different real-time contextual domains. Also, we present ideas on how to plug the data from a third-party application into the proposed architecture.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129332916","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Despite the fact that the problem of pathfinding exists for quite a while, most of the related work is focused on methods that are applicable only in 2D environment. However, there is a demand for methods that can provide a solution for pathfinding tasks in 3D working space. These tasks mostly come from area of robotics and game development, where it is often required to get a solution in real-time. The main problem, besides the lack of appropriate methods, is that in 3D environment the working space size increases greatly in comparison with 2D case, and it becomes harder to construct a precise representation of this space while maintaining low size of search graph. To overcome this, methods that lower graph size, whether explicitly or implicitly, are used. This paper provides a brief review on some of existing approaches, comparison of their effectiveness and presentation a new method of real-time pathfinding in 3D environment that can be helpful in real-time calculations (i.e. game development area). This brief review will be referred to when comparing proposed method with existing ones. A proposed method combines advantages of using octree structure as search graph (such as sparce free space representation) and hierarchical path planning (effective reduction of graph size by clustering). Thus, an appreciable speed-up is achieved.
{"title":"Octree-Based Hierarchical 3D Pathfinding Optimization of Three-Dimensional Pathfinding","authors":"Timur Muratov, A. Zagarskikh","doi":"10.1145/3386164.3386181","DOIUrl":"https://doi.org/10.1145/3386164.3386181","url":null,"abstract":"Despite the fact that the problem of pathfinding exists for quite a while, most of the related work is focused on methods that are applicable only in 2D environment. However, there is a demand for methods that can provide a solution for pathfinding tasks in 3D working space. These tasks mostly come from area of robotics and game development, where it is often required to get a solution in real-time. The main problem, besides the lack of appropriate methods, is that in 3D environment the working space size increases greatly in comparison with 2D case, and it becomes harder to construct a precise representation of this space while maintaining low size of search graph. To overcome this, methods that lower graph size, whether explicitly or implicitly, are used. This paper provides a brief review on some of existing approaches, comparison of their effectiveness and presentation a new method of real-time pathfinding in 3D environment that can be helpful in real-time calculations (i.e. game development area). This brief review will be referred to when comparing proposed method with existing ones. A proposed method combines advantages of using octree structure as search graph (such as sparce free space representation) and hierarchical path planning (effective reduction of graph size by clustering). Thus, an appreciable speed-up is achieved.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130544307","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Following the trends of electrification, the energy storage of vehicles is gaining importance as the most expensive part of an electric car. Since lithium-ion batteries are perishable goods and underlie e. g. aging effects, environmental and operating conditions during manufacturing and car usage need close supervision. With regard to the paradigm of digital twins, data from various life cycle phases needs to be collected and processed to improve the general quality of the system. To achieve this complex task, a suitable framework is needed in order to operate the fleet of digital twins during manufacturing processes, the automotive usage and a potential second life. Based on a literature review, we formulate requirements for a digital twin framework in the field of battery systems. We propose a framework to develop and operate a fleet of digital twins during all life cycle phases. Results feature a case study in which we implement the stated framework in a cloud-computing environment using early stages of battery system production as test a bed. With the help of a self-discharge model of li-ion cells, the system can estimate the SOC of battery modules and provide this information to the arrival testing procedures.
{"title":"Cloud-Based Battery Digital Twin Middleware Using Model-Based Development","authors":"Lukas Merkle","doi":"10.1145/3386164.3387296","DOIUrl":"https://doi.org/10.1145/3386164.3387296","url":null,"abstract":"Following the trends of electrification, the energy storage of vehicles is gaining importance as the most expensive part of an electric car. Since lithium-ion batteries are perishable goods and underlie e. g. aging effects, environmental and operating conditions during manufacturing and car usage need close supervision. With regard to the paradigm of digital twins, data from various life cycle phases needs to be collected and processed to improve the general quality of the system. To achieve this complex task, a suitable framework is needed in order to operate the fleet of digital twins during manufacturing processes, the automotive usage and a potential second life. Based on a literature review, we formulate requirements for a digital twin framework in the field of battery systems. We propose a framework to develop and operate a fleet of digital twins during all life cycle phases. Results feature a case study in which we implement the stated framework in a cloud-computing environment using early stages of battery system production as test a bed. With the help of a self-discharge model of li-ion cells, the system can estimate the SOC of battery modules and provide this information to the arrival testing procedures.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"482 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124496218","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Email has become a necessity for this new generation for official communication purposes. As the use of Internet is becoming more and more the risk of being caught into its darker side is so common. The major concern is spam, which is growing exponentially, and the users are becoming victim of it on daily basis. This paper proposes a hybrid machine learning classification model for the spam classification on the spambase dataset. This model uses the four classification algorithms namely Ensemble Classification, Decision Tree, Random Forest and Support Vector Machine (SVM). There are two phases; First phase deals with the classification of spambase dataset in two classes i.e. spam and ham with Decision Tree machine learning algorithm and the second phase comprises of classification improvisation of the output produced by phase one with four machine learning algorithms i.e. Decision Tree, Random Forest, Support Vector Machine (SVM) and Ensemble Learning. The experiment shows a very promising result with improvised accuracy in second phase.
{"title":"Machine Learning Techniques for Classification of Spambase Dataset: A Hybrid Approach","authors":"Shikha Verma, A. Gautam","doi":"10.1145/3386164.3389089","DOIUrl":"https://doi.org/10.1145/3386164.3389089","url":null,"abstract":"Email has become a necessity for this new generation for official communication purposes. As the use of Internet is becoming more and more the risk of being caught into its darker side is so common. The major concern is spam, which is growing exponentially, and the users are becoming victim of it on daily basis. This paper proposes a hybrid machine learning classification model for the spam classification on the spambase dataset. This model uses the four classification algorithms namely Ensemble Classification, Decision Tree, Random Forest and Support Vector Machine (SVM). There are two phases; First phase deals with the classification of spambase dataset in two classes i.e. spam and ham with Decision Tree machine learning algorithm and the second phase comprises of classification improvisation of the output produced by phase one with four machine learning algorithms i.e. Decision Tree, Random Forest, Support Vector Machine (SVM) and Ensemble Learning. The experiment shows a very promising result with improvised accuracy in second phase.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"78 8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128102761","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Edge-cloud computing provides performance guarantees for IoT applications which are real-time or security sensitive. The new placement of edge-cloud services leverages resources both in Cloud Data Centers (CDC) and at the edge of the network. A computation task can be divided into subtasks and offloaded to different edge/cloud servers, which are donated as offloading destinations. Offloading destination heterogeneity and different architecture of Edge Data Center (EDC) and CDC bring challenges to computation offloading. One critical issue in edge-cloud computing is energy consumption in computation offloading. The existing computation offloading strategies either ignored energy consumption or ignored delay and/or security constraints. Meta-heuristic strategies have been used widely to design heuristic resource allocation algorithms in CDC. This paper aims to explore meta-heuristic energy-efficient computation offloading (EE-CO) approaches with the objective to meet the delay and security constraints, while minimizing energy consumption. To achieve the goal, we investigated the performance of the Ant-Colony-Optimization (ACO) strategies combining with mixed integer programming (MIP). We propose an ACO-based computation offloading strategy, which including two algorithms, called EA-OMIP and EA-RMIP, respectively. The only difference of them is the construction method of integer programming models. Simulations are carried out to value the performance of proposed two algorithms. We also give an analysis of the experimental results in terms of the subtask acceptance ratio, revenue of the cloud service provider (CSP), and the resource utilization.
{"title":"A Meta-Heuristic Computation Offloading Strategy for IoT Applications in an Edge-Cloud Framework","authors":"Xuezhen Huang, Yang Yang, Xinglu Wu","doi":"10.1145/3386164.3390513","DOIUrl":"https://doi.org/10.1145/3386164.3390513","url":null,"abstract":"Edge-cloud computing provides performance guarantees for IoT applications which are real-time or security sensitive. The new placement of edge-cloud services leverages resources both in Cloud Data Centers (CDC) and at the edge of the network. A computation task can be divided into subtasks and offloaded to different edge/cloud servers, which are donated as offloading destinations. Offloading destination heterogeneity and different architecture of Edge Data Center (EDC) and CDC bring challenges to computation offloading. One critical issue in edge-cloud computing is energy consumption in computation offloading. The existing computation offloading strategies either ignored energy consumption or ignored delay and/or security constraints. Meta-heuristic strategies have been used widely to design heuristic resource allocation algorithms in CDC. This paper aims to explore meta-heuristic energy-efficient computation offloading (EE-CO) approaches with the objective to meet the delay and security constraints, while minimizing energy consumption. To achieve the goal, we investigated the performance of the Ant-Colony-Optimization (ACO) strategies combining with mixed integer programming (MIP). We propose an ACO-based computation offloading strategy, which including two algorithms, called EA-OMIP and EA-RMIP, respectively. The only difference of them is the construction method of integer programming models. Simulations are carried out to value the performance of proposed two algorithms. We also give an analysis of the experimental results in terms of the subtask acceptance ratio, revenue of the cloud service provider (CSP), and the resource utilization.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130763723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Autonomous manipulator grasp is an important issue in robotics research. To obtain the optimal grasp pose, we combine manipulator vision and deep learning to realize the artificial intelligence of the manipulator grasp. We adopt the idea of using residual network to improve the generative grasping convolutional neural network (GG-CNN). Firstly, we build a convolution residual module. By piling multi-layer of residual modules, we can build the residual network and deepen the depth of the convolutional neural network, which is used as the main part to improve GG-CNN. Improved GG-CNN based on deep residual network enhances the accuracy of the optimal grasping pose generation of the manipulator. Experimental results show that the accuracy of the improved GG-CNN model based on residual network reaches 88%, which is much higher than the original model's accuracy of 72%. It significantly improves the accuracy of the model to predict the optimal grasp pose of the manipulator.
{"title":"Robotic Grasp Pose Estimation Oriented Deep Learning Algorithm Based on Residual Network","authors":"Fan Bai, Renjie Yao, Maoning Chen, Zhexin Cui","doi":"10.1145/3386164.3389081","DOIUrl":"https://doi.org/10.1145/3386164.3389081","url":null,"abstract":"Autonomous manipulator grasp is an important issue in robotics research. To obtain the optimal grasp pose, we combine manipulator vision and deep learning to realize the artificial intelligence of the manipulator grasp. We adopt the idea of using residual network to improve the generative grasping convolutional neural network (GG-CNN). Firstly, we build a convolution residual module. By piling multi-layer of residual modules, we can build the residual network and deepen the depth of the convolutional neural network, which is used as the main part to improve GG-CNN. Improved GG-CNN based on deep residual network enhances the accuracy of the optimal grasping pose generation of the manipulator. Experimental results show that the accuracy of the improved GG-CNN model based on residual network reaches 88%, which is much higher than the original model's accuracy of 72%. It significantly improves the accuracy of the model to predict the optimal grasp pose of the manipulator.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130820369","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Face recognition have come from considering various aspects of this specialized perception problem such as apply for help checking attendant. In order to solve this problem, many systems have been completely changed due to this evolve to achieve more accurate results. This research aims to develop the facing attendant system to be more effective and the mechanic of the system, which students can easily verify. The cloud storage was used for Attendance System. The experiment of this research is to find the way to recognize the face by using the technique of Neural Networks, which can correctly recognize up to 95%. This model can apply with school and university.
{"title":"Face Recognition for Attendance System Using Neural Networks Technique","authors":"D. Prangchumpol","doi":"10.1145/3386164.3389102","DOIUrl":"https://doi.org/10.1145/3386164.3389102","url":null,"abstract":"Face recognition have come from considering various aspects of this specialized perception problem such as apply for help checking attendant. In order to solve this problem, many systems have been completely changed due to this evolve to achieve more accurate results. This research aims to develop the facing attendant system to be more effective and the mechanic of the system, which students can easily verify. The cloud storage was used for Attendance System. The experiment of this research is to find the way to recognize the face by using the technique of Neural Networks, which can correctly recognize up to 95%. This model can apply with school and university.","PeriodicalId":231209,"journal":{"name":"Proceedings of the 2019 3rd International Symposium on Computer Science and Intelligent Control","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131119411","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：