∗Computation offloading is currently future-oriented, which has not been large-range deployed. However, it is a useful tool for the growing computing requirements for mobile devices. Now trading apps, such as TradingView and Futu, tend to provide either the full functionality to run real-time scripts like variants of technical, or autonomous trading strategies, turning out to increase computation scale dramatically or providing just limited functionalities. Current solutions either degrade responsibility of the mobile devices or use cloud computing, which produces more latency compared to using 5GMobile Edge Computing (MEC) units. This paper proposes a novel comparison of computing locally (or on MEC units) and a method to evaluate the offloaded acceleration rate. The result shows the suitable measure to offload computation to MEC units. In addition, it also shows that it is possible to process real-time scripts on the fog layer in some situations. It can be concluded that the proposed method reduces the latency of the whole trading system.
{"title":"Computation Offloading for Better Real-Time Technical Market Analysis on Mobile Devices","authors":"Gufeng Shen","doi":"10.1145/3469951.3469964","DOIUrl":"https://doi.org/10.1145/3469951.3469964","url":null,"abstract":"∗Computation offloading is currently future-oriented, which has not been large-range deployed. However, it is a useful tool for the growing computing requirements for mobile devices. Now trading apps, such as TradingView and Futu, tend to provide either the full functionality to run real-time scripts like variants of technical, or autonomous trading strategies, turning out to increase computation scale dramatically or providing just limited functionalities. Current solutions either degrade responsibility of the mobile devices or use cloud computing, which produces more latency compared to using 5GMobile Edge Computing (MEC) units. This paper proposes a novel comparison of computing locally (or on MEC units) and a method to evaluate the offloaded acceleration rate. The result shows the suitable measure to offload computation to MEC units. In addition, it also shows that it is possible to process real-time scripts on the fog layer in some situations. It can be concluded that the proposed method reduces the latency of the whole trading system.","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"41 4","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114093128","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}
∗With the continuous development of Unmanned Aerial Vehicle (UAV) technology and its industry, the detection and recognition technology of UAV have attracted the attention of researchers. In this paper, the author focuses on the defects and deficiencies of traditional radar, visual and acoustic UAV detection technology. Considering that the UAV’s own radio communication signal can be used for detection, a UAV signal classification method based on deep learning is proposed. This algorithm can extract the characteristics of UAV Communication Law, so as to achieve the target classification. The experimental results show that the average recognition rate of UAV is 95% in the test, and the recognition rate of most types of UAVs is more than 98%. In addition, the classification rate for the flight attitudes of UAVs can reach more than 95%. Therefore, it can be concluded that the classification algorithm designed in this paper can effectively meet the needs of UAV detection and recognition in the actual scene.
{"title":"Research on UAV Signal Classification Algorithm Based on Deep Learning","authors":"Yunsong Zhao","doi":"10.1145/3469951.3469956","DOIUrl":"https://doi.org/10.1145/3469951.3469956","url":null,"abstract":"∗With the continuous development of Unmanned Aerial Vehicle (UAV) technology and its industry, the detection and recognition technology of UAV have attracted the attention of researchers. In this paper, the author focuses on the defects and deficiencies of traditional radar, visual and acoustic UAV detection technology. Considering that the UAV’s own radio communication signal can be used for detection, a UAV signal classification method based on deep learning is proposed. This algorithm can extract the characteristics of UAV Communication Law, so as to achieve the target classification. The experimental results show that the average recognition rate of UAV is 95% in the test, and the recognition rate of most types of UAVs is more than 98%. In addition, the classification rate for the flight attitudes of UAVs can reach more than 95%. Therefore, it can be concluded that the classification algorithm designed in this paper can effectively meet the needs of UAV detection and recognition in the actual scene.","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115443710","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}
With the rapid development of E-commerce, more and more users are buying clothes through the Internet, and "image search" for clothing images has become a popular research direction. The current "image search" technology mainly relies on the results of feature extraction of the whole image, but cannot focus on the parts of the clothing, and the background of the clothing image is generally complex, resulting in low accuracy of clothing image retrieval, so we propose a retrieval method based on clothing image detection and segmentation. Firstly, Mask R-CNN is used to detect and segment the image to get the information of garment body, collar parts, sleeve category and pocket positions, then VGG16 is used to extract 512-dimensional features from the garment body and collar parts, based on this information, the similarity between the garment to be retrieved and the garment in the database is calculated one by one. We calculate the similarity by weighting the cosine similarity of 512-dimensional features of the garment body and collar, as well as the similarity of the sleeves and pockets. The search results are presented to the user according to the descending order of similarity. The experimental results show that the method can focus on the whole garment as well as their parts, thus enabling retrieval based on garment style. It also allows users to adjust the weights of each part and can return the search results that best meet their individual needs
{"title":"Clothing Image Retrieval Based on Parts Detection and Segmentation","authors":"Qiubo Huang, X. Han, Ting Lu, Guohua Liu","doi":"10.1145/3469951.3469961","DOIUrl":"https://doi.org/10.1145/3469951.3469961","url":null,"abstract":"With the rapid development of E-commerce, more and more users are buying clothes through the Internet, and \"image search\" for clothing images has become a popular research direction. The current \"image search\" technology mainly relies on the results of feature extraction of the whole image, but cannot focus on the parts of the clothing, and the background of the clothing image is generally complex, resulting in low accuracy of clothing image retrieval, so we propose a retrieval method based on clothing image detection and segmentation. Firstly, Mask R-CNN is used to detect and segment the image to get the information of garment body, collar parts, sleeve category and pocket positions, then VGG16 is used to extract 512-dimensional features from the garment body and collar parts, based on this information, the similarity between the garment to be retrieved and the garment in the database is calculated one by one. We calculate the similarity by weighting the cosine similarity of 512-dimensional features of the garment body and collar, as well as the similarity of the sleeves and pockets. The search results are presented to the user according to the descending order of similarity. The experimental results show that the method can focus on the whole garment as well as their parts, thus enabling retrieval based on garment style. It also allows users to adjust the weights of each part and can return the search results that best meet their individual needs","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"91 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123967839","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}
Nguyen Pham Nguyen Xuan, Tham Tran Thi, Thang DO Minh, Duy Tran Ngoc Bao
In the quantitative assessment on the success of treatment, the automatic detection of acne pixels from digital color images would be helpful. In this paper, we proposed an automatic acne detection method through the processing of facial images taken by the smartphone based on the image processing. In this approach, the RGB image is transformed into various color spaces based on the differences between features of each acne lesion type. This method has been used the a* channel of the CIELab color space to detect the inflammatory acne (papules and pustules). The S channel of HSV color space was used to detect the non-inflammatory acne (whiteheads and blackheads). A multi-level threshold is then used to make acne extraction and blob detection. The effectiveness of the proposed procedure is shown by experimental results. We showed the possibility of detecting 4 types of acne lesions (whiteheads, blackheads, papules, pustules) with different skin colors and different smartphones in this experiment by applying a combination of several color spaces. The result shows a recall of about 85.71% in detecting different acne types at a reasonable processing time. This is the remise to help doctors to assess the level of acne on the patient's face in an effective and time-saving way.
{"title":"A Multilevel Thresholding Approach for Acne Detection in Medical Treatment","authors":"Nguyen Pham Nguyen Xuan, Tham Tran Thi, Thang DO Minh, Duy Tran Ngoc Bao","doi":"10.1145/3469951.3469955","DOIUrl":"https://doi.org/10.1145/3469951.3469955","url":null,"abstract":"In the quantitative assessment on the success of treatment, the automatic detection of acne pixels from digital color images would be helpful. In this paper, we proposed an automatic acne detection method through the processing of facial images taken by the smartphone based on the image processing. In this approach, the RGB image is transformed into various color spaces based on the differences between features of each acne lesion type. This method has been used the a* channel of the CIELab color space to detect the inflammatory acne (papules and pustules). The S channel of HSV color space was used to detect the non-inflammatory acne (whiteheads and blackheads). A multi-level threshold is then used to make acne extraction and blob detection. The effectiveness of the proposed procedure is shown by experimental results. We showed the possibility of detecting 4 types of acne lesions (whiteheads, blackheads, papules, pustules) with different skin colors and different smartphones in this experiment by applying a combination of several color spaces. The result shows a recall of about 85.71% in detecting different acne types at a reasonable processing time. This is the remise to help doctors to assess the level of acne on the patient's face in an effective and time-saving way.","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"138 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125743233","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}
Yashar Deldjoo, Tommaso Di Noia, Eugenio Di Sciascio, Gaetano Pernisco, V. Renó, E. Stella
Most autonomous driving applications leverage RGB images representing the surrounding environment that contain useful appearance features but with a cost in terms of geometric features. On the other side, 3D point clouds generated by LIDAR sensors can provide more geometric 3D information with high accuracy and robustness but with a loss on appearance features. Regardless of the adopted technology, object tracking in autonomous driving scenarios suffers from the so-called error drift in detecting objects over time/frames. This work investigates the car tracking problem in an urban scenario, leveraging 3D point clouds. In particular, we have set our goal to mitigate the typical error drift that characterizes the classic tracking algorithm and, to this aim, proposed a system able to reduce the drift error by detection. An extensive experimental evaluation on the KITTI dataset shows the improvement in our solution's performance compared to state-of-the-art approaches.
{"title":"Towards Improving Car Point-Cloud Tracking Via Detection Updates","authors":"Yashar Deldjoo, Tommaso Di Noia, Eugenio Di Sciascio, Gaetano Pernisco, V. Renó, E. Stella","doi":"10.1145/3469951.3469957","DOIUrl":"https://doi.org/10.1145/3469951.3469957","url":null,"abstract":"Most autonomous driving applications leverage RGB images representing the surrounding environment that contain useful appearance features but with a cost in terms of geometric features. On the other side, 3D point clouds generated by LIDAR sensors can provide more geometric 3D information with high accuracy and robustness but with a loss on appearance features. Regardless of the adopted technology, object tracking in autonomous driving scenarios suffers from the so-called error drift in detecting objects over time/frames. This work investigates the car tracking problem in an urban scenario, leveraging 3D point clouds. In particular, we have set our goal to mitigate the typical error drift that characterizes the classic tracking algorithm and, to this aim, proposed a system able to reduce the drift error by detection. An extensive experimental evaluation on the KITTI dataset shows the improvement in our solution's performance compared to state-of-the-art approaches.","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123003926","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}
Saifullah Tumrani, Parivish Parivish, A. Khan, Wazir Ali
Vehicle Re-Identification is the task of finding images of the same vehicle with different views across a surveillance camera network, which is a very beneficial yet challenging task. Huge intra-class differences and small inter-class difference makes this task hard to tackle. Appearance-based information is utilized in this paper to cope with vehicle re-identification problem; we have proposed a deep learning technique by incorporating poses of vehicles generated by pose estimation network and visual information. When query image is given, the two-stream network generates a feature embedding by concatenating pose feature from pose network. Extensive experiments are done on two of the benchmark datasets of vehicle re-identification VeRi-776 and VehicleID. Experimental results are supporting the competitiveness of the proposed method with recent state-of-the-art methods.
{"title":"Two Stream Pose Guided Network for Vehicle Re-identification","authors":"Saifullah Tumrani, Parivish Parivish, A. Khan, Wazir Ali","doi":"10.1145/3469951.3469954","DOIUrl":"https://doi.org/10.1145/3469951.3469954","url":null,"abstract":"Vehicle Re-Identification is the task of finding images of the same vehicle with different views across a surveillance camera network, which is a very beneficial yet challenging task. Huge intra-class differences and small inter-class difference makes this task hard to tackle. Appearance-based information is utilized in this paper to cope with vehicle re-identification problem; we have proposed a deep learning technique by incorporating poses of vehicles generated by pose estimation network and visual information. When query image is given, the two-stream network generates a feature embedding by concatenating pose feature from pose network. Extensive experiments are done on two of the benchmark datasets of vehicle re-identification VeRi-776 and VehicleID. Experimental results are supporting the competitiveness of the proposed method with recent state-of-the-art methods.","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123215852","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}
Histopathology serves as the gold standard for tumor diagnosis. Whole slide scanners have made computer vision-based methods available for pathologists to locate regions of high diagnostic significance. An essential step of whole slide image (WSI) diagnosis is the segmentation of the tumor region by generating a tumor probability heatmap. Most WSI diagnosis methods use patch-based classifiers or segmentation models, they both require a large set of training patches from annotated WSIs. Annotating WSIs is time-consuming and laborious. Active learning is a method that can suggest the most informative unlabeled data for annotation, but traditional active learning methods are not directly applicable for WSIs. Meanwhile, unannotated WSIs also contain rich information that can be further exploited by self-supervised learning. By utilizing unannotated data alongside active learning, we proposed a self-supervised active learning framework for tumor region segmentation of WSIs. The proposed method is evaluated on the public available CAMELYON dataset and achieved satisfying performance using 3% of the annotated data.
{"title":"Reducing the Annotation Cost of Whole Slide Histology Images using Active Learning","authors":"Xu Jin, Hong An, Jue Wang, Ke Wen, Zheng Wu","doi":"10.1145/3469951.3469960","DOIUrl":"https://doi.org/10.1145/3469951.3469960","url":null,"abstract":"Histopathology serves as the gold standard for tumor diagnosis. Whole slide scanners have made computer vision-based methods available for pathologists to locate regions of high diagnostic significance. An essential step of whole slide image (WSI) diagnosis is the segmentation of the tumor region by generating a tumor probability heatmap. Most WSI diagnosis methods use patch-based classifiers or segmentation models, they both require a large set of training patches from annotated WSIs. Annotating WSIs is time-consuming and laborious. Active learning is a method that can suggest the most informative unlabeled data for annotation, but traditional active learning methods are not directly applicable for WSIs. Meanwhile, unannotated WSIs also contain rich information that can be further exploited by self-supervised learning. By utilizing unannotated data alongside active learning, we proposed a self-supervised active learning framework for tumor region segmentation of WSIs. The proposed method is evaluated on the public available CAMELYON dataset and achieved satisfying performance using 3% of the annotated data.","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123294822","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}
T. Shen, D. Wang, Kayton Wai Keung Cheung, M. C. Chan, King Hung Chiu, Yiu Kei Li
Face detection and gender classification by Deep Neural Networks can find application in areas such as video surveillance, customized advertisement, and human-computer interaction. This paper presents a real-time single-shot multi-face gender detector based on Convolutional neural network (CNN). The proposed method not only detects face but also classifies the gender of persons in the wild, meaning in images with a high variability in pose, illumination, and occlusion. To train and evaluate the results, a new annotated set of face images is created. Our experimental results show that the proposed method achieves excellent performance in term of speed and accuracy.
{"title":"A Real-Time Single-Shot Multi-Face Detection, Landmark Localization, and Gender Classification","authors":"T. Shen, D. Wang, Kayton Wai Keung Cheung, M. C. Chan, King Hung Chiu, Yiu Kei Li","doi":"10.1145/3469951.3469952","DOIUrl":"https://doi.org/10.1145/3469951.3469952","url":null,"abstract":"Face detection and gender classification by Deep Neural Networks can find application in areas such as video surveillance, customized advertisement, and human-computer interaction. This paper presents a real-time single-shot multi-face gender detector based on Convolutional neural network (CNN). The proposed method not only detects face but also classifies the gender of persons in the wild, meaning in images with a high variability in pose, illumination, and occlusion. To train and evaluate the results, a new annotated set of face images is created. Our experimental results show that the proposed method achieves excellent performance in term of speed and accuracy.","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116319306","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 recent years, telecom operators and large companies are eager to obtain value from the edge of the network, and the priority of cloud computing has been transferred from the center to the edge. In addition, with the comprehensive deployment of 5G base station (BS), the number of 5G users has been largely increased. For 5G users, they expect to have a better experience of high bandwidth and low latency. Thus, the Mobile Edge Computing (MEC) came into being. MEC brings the capability from the center to the edge of the mobile network. Requests and data of User equipment (UE) has been underlined in MEC. These requests and data will be analyzed and disposed at the edge without being uploaded to the cloud center, which diminishes the latency efficiently. Besides, with the help of machine learning, MEC can show a better performance. This paper is aimed at studying superiorities of MEC itself and integration of machine learning with MEC, and intelligent applications they will bring. This paper first discusses the concept and architecture of MEC, then the advantages of MEC are listed. Next, the improvements of integration of machine learning with MEC and the intelligent applications which employ these technologies will be introduced. Finally, the deficiencies and future research trend of MEC will be discussed. After that, conclusion can be drought that MEC augment the performance of speed, security and privacy, energy saving and reliability. Furthermore, integration of machine learning with MEC can provide better resource management and offloading decision.
{"title":"Integration of Machine Learning with MEC for Intelligent Applications","authors":"Zhou Ye","doi":"10.1145/3469951.3469966","DOIUrl":"https://doi.org/10.1145/3469951.3469966","url":null,"abstract":"∗In recent years, telecom operators and large companies are eager to obtain value from the edge of the network, and the priority of cloud computing has been transferred from the center to the edge. In addition, with the comprehensive deployment of 5G base station (BS), the number of 5G users has been largely increased. For 5G users, they expect to have a better experience of high bandwidth and low latency. Thus, the Mobile Edge Computing (MEC) came into being. MEC brings the capability from the center to the edge of the mobile network. Requests and data of User equipment (UE) has been underlined in MEC. These requests and data will be analyzed and disposed at the edge without being uploaded to the cloud center, which diminishes the latency efficiently. Besides, with the help of machine learning, MEC can show a better performance. This paper is aimed at studying superiorities of MEC itself and integration of machine learning with MEC, and intelligent applications they will bring. This paper first discusses the concept and architecture of MEC, then the advantages of MEC are listed. Next, the improvements of integration of machine learning with MEC and the intelligent applications which employ these technologies will be introduced. Finally, the deficiencies and future research trend of MEC will be discussed. After that, conclusion can be drought that MEC augment the performance of speed, security and privacy, energy saving and reliability. Furthermore, integration of machine learning with MEC can provide better resource management and offloading decision.","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115837346","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}
Cardiac arrhythmia is a common sign of heart disease. In modern society, heart disease is always one of the main diseases threatening human health. Medical instruments collect related attributes to make better diagnosis prediction of the disease. This paper applies different feature selection methods including filters and wrappers combining with machine learning methods (SVM, Naive Bayes, Random Forest, C4.5) on the arrhythmia dataset to compare their performances. Results show that filters and wrappers perform both well while filters cost less time. Among them, random forest with the wrapper method has the highest accuracy.
{"title":"Comparison of Feature Selection Methods on Arrhythmia Dataset","authors":"Liu Ziheng","doi":"10.1145/3469951.3469963","DOIUrl":"https://doi.org/10.1145/3469951.3469963","url":null,"abstract":"Cardiac arrhythmia is a common sign of heart disease. In modern society, heart disease is always one of the main diseases threatening human health. Medical instruments collect related attributes to make better diagnosis prediction of the disease. This paper applies different feature selection methods including filters and wrappers combining with machine learning methods (SVM, Naive Bayes, Random Forest, C4.5) on the arrhythmia dataset to compare their performances. Results show that filters and wrappers perform both well while filters cost less time. Among them, random forest with the wrapper method has the highest accuracy.","PeriodicalId":313453,"journal":{"name":"Proceedings of the 2021 3rd International Conference on Image Processing and Machine Vision","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123342859","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：