Pub Date : 2022-03-01DOI: 10.1109/ICICT55905.2022.00011
Oliver Simonoski, Izabela Mitreska, D. C. Bogatinoska, A. Hristov
The novel coronavirus (COVID-19) that was first reported at the end of 2019 has impacted almost every aspect of our lives. While healthcare workers fight the virus in the front line, we do our part by creating an Electronic Health Records system that tracks state and federal prisoner's data through the countries in order to monitor COVID-19 cases in the U.S. The main objective of our system is to visualize the relationship between current data of deaths per day in both, state and federal prisons in U.S. State. In order to accomplish this process, we combine COVID-19 case and death rates into a single data collection which will show that state prisons have consistently reported greater rates of COVID-19 than federal prisons. The obtained results from the process satisfied our expectations and provide the efficiency of implementing Electronic Health Records Systems.
{"title":"EHR system tracks prisoner's data across countries to monitor COVID-19 in U.S. State and Federal Prisons","authors":"Oliver Simonoski, Izabela Mitreska, D. C. Bogatinoska, A. Hristov","doi":"10.1109/ICICT55905.2022.00011","DOIUrl":"https://doi.org/10.1109/ICICT55905.2022.00011","url":null,"abstract":"The novel coronavirus (COVID-19) that was first reported at the end of 2019 has impacted almost every aspect of our lives. While healthcare workers fight the virus in the front line, we do our part by creating an Electronic Health Records system that tracks state and federal prisoner's data through the countries in order to monitor COVID-19 cases in the U.S. The main objective of our system is to visualize the relationship between current data of deaths per day in both, state and federal prisons in U.S. State. In order to accomplish this process, we combine COVID-19 case and death rates into a single data collection which will show that state prisons have consistently reported greater rates of COVID-19 than federal prisons. The obtained results from the process satisfied our expectations and provide the efficiency of implementing Electronic Health Records Systems.","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125552251","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}
Pub Date : 2022-03-01DOI: 10.1109/ICICT55905.2022.00020
V. O. Nancy, Meenakshi S. Arya, N. Nitin
The known peculiar type of cancer type is melanoma. It arises as pigment and is hard to find in the initial stages. The persistence level is 99% when identified in the early stage. Classification and identification of malignant tumors in skin lesions are crucial. The main goal is to sort the lesion images to seven important classes and identify the cancerous and non-cancerous tumors at the earliest using deep learning techniques. The efficient way for deep learning outcomes is to use a large volume and high-quality training dataset. Existing datasets are effectively not sufficient for training the model. The techniques for data augmentation are effective ways to build highly accurate classifiers from insufficient data. The proposed methodology offered the effective strategy for diagnosing the malignant tumor is a CNN-based model. CNN is specifically used to recognize and classify images. The framework is trained with data that has been labeled with the appropriate class. A similar framework has been trained with augmented and non-augmented lesion images for knowing the malignant lesions. The results are compared to both original data and augmented data. The model evaluated, the accuracy occurred for augmented data is 97.86%.
{"title":"Impact of Data Augmentation on Skin Lesion Classification Using Deep Learning","authors":"V. O. Nancy, Meenakshi S. Arya, N. Nitin","doi":"10.1109/ICICT55905.2022.00020","DOIUrl":"https://doi.org/10.1109/ICICT55905.2022.00020","url":null,"abstract":"The known peculiar type of cancer type is melanoma. It arises as pigment and is hard to find in the initial stages. The persistence level is 99% when identified in the early stage. Classification and identification of malignant tumors in skin lesions are crucial. The main goal is to sort the lesion images to seven important classes and identify the cancerous and non-cancerous tumors at the earliest using deep learning techniques. The efficient way for deep learning outcomes is to use a large volume and high-quality training dataset. Existing datasets are effectively not sufficient for training the model. The techniques for data augmentation are effective ways to build highly accurate classifiers from insufficient data. The proposed methodology offered the effective strategy for diagnosing the malignant tumor is a CNN-based model. CNN is specifically used to recognize and classify images. The framework is trained with data that has been labeled with the appropriate class. A similar framework has been trained with augmented and non-augmented lesion images for knowing the malignant lesions. The results are compared to both original data and augmented data. The model evaluated, the accuracy occurred for augmented data is 97.86%.","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132584616","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}
Pub Date : 2022-03-01DOI: 10.1109/ICICT55905.2022.00046
Zhenyi Yang, Rebecca Miao, Marina Orlova, I. Nechepurenko, V. Gavrishchaka
Early detection of developing abnormalities or treatment effects could critically enhance success of prevention and treatment strategies. While many advanced technologies are available for accurate clinical diagnostics, their wide 24/7 usage required for early preventive alerts including detection of emerging intermittent patterns is not feasible. Although modern wearable devices offer affordable continuous recording of physiological data, data collected over long-term necessarily have significantly lower resolution due to technological limitations leading to sharp accuracy deterioration of mainstream diagnostic techniques. Recently, we demonstrated that some of these challenges can be resolved by hybrid framework where boosting algorithms are used for enhancement of existing domain-expert models with further non-linear combination of boosted ensemble components via deep learning or other machine learning algorithms. While normal-abnormal differentiation performance of such hybrid indicators was confirmed using real cardio data from www.physionet.org, their applicability to more challenging problem of early-stage detection of emerging abnormalities or treatment effects remain unknown since long-term transition data from normal to abnormal states is not available. Here we propose a framework for verification and enhancement of indicator abilities for such early detection using simulated transition paths obtained by sampling real normal/abnormal data and employing realistic synthetic data generated by physics-based models. Robust performance of our hybrid indicators was confirmed in cases where other existing approaches fail.
{"title":"Discovery of early-alert indicators using hybrid ensemble learning and generative physics-based models","authors":"Zhenyi Yang, Rebecca Miao, Marina Orlova, I. Nechepurenko, V. Gavrishchaka","doi":"10.1109/ICICT55905.2022.00046","DOIUrl":"https://doi.org/10.1109/ICICT55905.2022.00046","url":null,"abstract":"Early detection of developing abnormalities or treatment effects could critically enhance success of prevention and treatment strategies. While many advanced technologies are available for accurate clinical diagnostics, their wide 24/7 usage required for early preventive alerts including detection of emerging intermittent patterns is not feasible. Although modern wearable devices offer affordable continuous recording of physiological data, data collected over long-term necessarily have significantly lower resolution due to technological limitations leading to sharp accuracy deterioration of mainstream diagnostic techniques. Recently, we demonstrated that some of these challenges can be resolved by hybrid framework where boosting algorithms are used for enhancement of existing domain-expert models with further non-linear combination of boosted ensemble components via deep learning or other machine learning algorithms. While normal-abnormal differentiation performance of such hybrid indicators was confirmed using real cardio data from www.physionet.org, their applicability to more challenging problem of early-stage detection of emerging abnormalities or treatment effects remain unknown since long-term transition data from normal to abnormal states is not available. Here we propose a framework for verification and enhancement of indicator abilities for such early detection using simulated transition paths obtained by sampling real normal/abnormal data and employing realistic synthetic data generated by physics-based models. Robust performance of our hybrid indicators was confirmed in cases where other existing approaches fail.","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127446908","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}
Pub Date : 2022-03-01DOI: 10.1109/ICICT55905.2022.00013
O. Uchida, Ryoji Yamaguchi, Kohei Cho
To collect and spread accurate information quickly is vital to minimize the damage caused by disasters. Then, the utilization of social media during disasters has been gaining attention. Based on such background, we developed a Twitter-based disaster information sharing system called DITS (disaster information tweeting system) & DIMS (disaster information mapping system) in previous studies. Using DITS, we can post and share disaster-related tweets with location information and the appropriate hashtags with simple operations. Furthermore, we can view the information posted using DITS on a map with DIMS. However, there is concern that if the system can only be used in disasters, it will not be adequately utilized when one occurs. The study then expanded DITS & DIMS into a system that can be used for more than just disasters, that is, to share daily local information, such as tourist and gourmet information. This study also changed the system to make it more usable.
迅速收集和传播准确的信息对于尽量减少灾害造成的损失至关重要。然后，在灾难中使用社交媒体已经引起了人们的关注。基于这样的背景，我们在之前的研究中开发了基于twitter的灾害信息共享系统DITS (disaster information tweeting system) & DIMS (disaster information mapping system)。使用DITS，我们可以通过简单的操作发布和共享带有位置信息和适当标签的与灾难相关的tweet。此外，我们可以在带有DIMS的地图上查看使用DITS发布的信息。但是，令人关切的是，如果该系统只能用于灾害，那么在灾害发生时就不能充分利用。随后，该研究将DITS & DIMS扩展为一个不仅仅用于灾难的系统，也就是说，可以共享当地的日常信息，如旅游和美食信息。这项研究还改变了系统，使其更可用。
{"title":"Extension of the Disaster Information Sharing System DITS & DIMS to a System Available on a Daily Use","authors":"O. Uchida, Ryoji Yamaguchi, Kohei Cho","doi":"10.1109/ICICT55905.2022.00013","DOIUrl":"https://doi.org/10.1109/ICICT55905.2022.00013","url":null,"abstract":"To collect and spread accurate information quickly is vital to minimize the damage caused by disasters. Then, the utilization of social media during disasters has been gaining attention. Based on such background, we developed a Twitter-based disaster information sharing system called DITS (disaster information tweeting system) & DIMS (disaster information mapping system) in previous studies. Using DITS, we can post and share disaster-related tweets with location information and the appropriate hashtags with simple operations. Furthermore, we can view the information posted using DITS on a map with DIMS. However, there is concern that if the system can only be used in disasters, it will not be adequately utilized when one occurs. The study then expanded DITS & DIMS into a system that can be used for more than just disasters, that is, to share daily local information, such as tourist and gourmet information. This study also changed the system to make it more usable.","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133059302","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}
Pub Date : 2022-03-01DOI: 10.1109/ICICT55905.2022.00019
B. Orazbayev, A. Zhumadillayeva, K. Dyussekeyev, K. Orazbayeva, T. Umarov, L. Kurmangaziyeva
The creation of intellectualized decision support systems (IDSS) for managing various, complex and difficult to formalize objects is currently a topical issue of science and practice. In this work, a subsystem for modeling ISSPR is created, which makes it possible to determine the effective operating modes of the catalytic reforming unit reactors. Mathematical models of the investigated reactors are developed on the basis of statistical data and fuzzy information obtained by expert assessment methods. Accordingly, the models for determining the volume of products produced from the output of the reactors are built in the form of statistical models, and for assessing the quality of the target product of the reforming process, i.e., catalyzate, based on expert information in the form of fuzzy models. An IDSS structure has been created to control the operating modes of the investigated reactors, which differs from the known structures of similar systems in that it includes a package of models of the control object and heuristic methods for solving decision-making problems, which make it possible to take into account the fuzziness of the initial information. In addition, the proposed IDSS structure contains a knowledge and data base, identifiers of model parameters, an intellectualized user interface and a block for explaining the selected solutions, which make it possible to increase the efficiency of the system. Describe the functions of the main functional blocks of the created IDSS.
{"title":"Modeling Subsystem for Optimizing Reforming Processes of an Intellectualized Decision Support System","authors":"B. Orazbayev, A. Zhumadillayeva, K. Dyussekeyev, K. Orazbayeva, T. Umarov, L. Kurmangaziyeva","doi":"10.1109/ICICT55905.2022.00019","DOIUrl":"https://doi.org/10.1109/ICICT55905.2022.00019","url":null,"abstract":"The creation of intellectualized decision support systems (IDSS) for managing various, complex and difficult to formalize objects is currently a topical issue of science and practice. In this work, a subsystem for modeling ISSPR is created, which makes it possible to determine the effective operating modes of the catalytic reforming unit reactors. Mathematical models of the investigated reactors are developed on the basis of statistical data and fuzzy information obtained by expert assessment methods. Accordingly, the models for determining the volume of products produced from the output of the reactors are built in the form of statistical models, and for assessing the quality of the target product of the reforming process, i.e., catalyzate, based on expert information in the form of fuzzy models. An IDSS structure has been created to control the operating modes of the investigated reactors, which differs from the known structures of similar systems in that it includes a package of models of the control object and heuristic methods for solving decision-making problems, which make it possible to take into account the fuzziness of the initial information. In addition, the proposed IDSS structure contains a knowledge and data base, identifiers of model parameters, an intellectualized user interface and a block for explaining the selected solutions, which make it possible to increase the efficiency of the system. Describe the functions of the main functional blocks of the created IDSS.","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130221598","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}
Pub Date : 2022-03-01DOI: 10.1109/icict55905.2022.00001
{"title":"Proceedings 2022 5th International Conference on Information and Computer Technologies ICICT 2022","authors":"","doi":"10.1109/icict55905.2022.00001","DOIUrl":"https://doi.org/10.1109/icict55905.2022.00001","url":null,"abstract":"","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127342955","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}
Pub Date : 2022-03-01DOI: 10.1109/ICICT55905.2022.00032
Toshiro Minami, Y. Ohura
The goal of the study presented in this paper is to understand attitudes and viewpoints of university students toward learning. We have been investigating text data obtained as answer texts for a term-end questionnaire in a class, and have found some interesting facts about students' attitude toward learning. This paper aims to investigate the texts further by using the part-of-speech (POS) information so that we can extract different features from those obtained in our former studies. In this paper, we develop a couple of distances between students so that we can see the features from different methods of measurement. As a result, we can find out some students who are different from other students regarding usage of POS.
{"title":"A Characterization of Word- Usage of Students Using Part-of-Speech Information","authors":"Toshiro Minami, Y. Ohura","doi":"10.1109/ICICT55905.2022.00032","DOIUrl":"https://doi.org/10.1109/ICICT55905.2022.00032","url":null,"abstract":"The goal of the study presented in this paper is to understand attitudes and viewpoints of university students toward learning. We have been investigating text data obtained as answer texts for a term-end questionnaire in a class, and have found some interesting facts about students' attitude toward learning. This paper aims to investigate the texts further by using the part-of-speech (POS) information so that we can extract different features from those obtained in our former studies. In this paper, we develop a couple of distances between students so that we can see the features from different methods of measurement. As a result, we can find out some students who are different from other students regarding usage of POS.","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121085172","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}
This paper proposes a spatial and channel attention mechanism module called SC-NET which is a lightweight yet effective method for deep convolutional neural networks. Recently, channel attention mechanism has been researched extensively and proved to be efficient in improvement of performance. However after carrying out rigorous empirical analysis, we find that channel attention and spatial channel attention improve the network's performance more efficiently. Therefore we incorporate both spatial information and cross-channel interaction in our SC-NET architecture. SC-NET is validated through extensive experiments on CASIA- WebFace and VGGFace2 datasets. By comparing our SC-NET with other methods, SC-NET has the best performance. Then when we apply our SC-NET to FaceNet(A Unified Embedding for Face Recognition and Clustering), FaceNet with SC-NET has achieved higher recognition accuracy than the original FaceNet and has reached state-of-the-art performance.
{"title":"SC-NET: Spatial and Channel Attention Mechanism for Enhancement in Face Recognition","authors":"Yefan Zhu, Yanhong Liang, Tang Kai, Kazushige Ouchi","doi":"10.1109/ICICT55905.2022.00036","DOIUrl":"https://doi.org/10.1109/ICICT55905.2022.00036","url":null,"abstract":"This paper proposes a spatial and channel attention mechanism module called SC-NET which is a lightweight yet effective method for deep convolutional neural networks. Recently, channel attention mechanism has been researched extensively and proved to be efficient in improvement of performance. However after carrying out rigorous empirical analysis, we find that channel attention and spatial channel attention improve the network's performance more efficiently. Therefore we incorporate both spatial information and cross-channel interaction in our SC-NET architecture. SC-NET is validated through extensive experiments on CASIA- WebFace and VGGFace2 datasets. By comparing our SC-NET with other methods, SC-NET has the best performance. Then when we apply our SC-NET to FaceNet(A Unified Embedding for Face Recognition and Clustering), FaceNet with SC-NET has achieved higher recognition accuracy than the original FaceNet and has reached state-of-the-art performance.","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"150 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131473374","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}
Pub Date : 2022-03-01DOI: 10.1109/ICICT55905.2022.00033
H. Pham
Computer operating systems are complex systems with many components which are difficult to monitor, analyze, and learn. Operating systems simulators can help to overcome these challenges. This paper provides a brief review of operating system simulators and proposes Web VizOS, a web-based framework for interactive simulations and visualizations of operating system main concepts and mechanisms. This framework can be used for: (i)development, research, and optimization of each individual operating system mechanism; (ii)comparison and analysis of alternative methods in the same category; and (iii) study and research of operating systems in whole. This framework is designed to be comprehensive and can include all fundamental mechanisms in process management, memory management, I/O control, file and disk management. This simulation framework is open and flexible enough that users can develop their own codes for operating system mechanisms independently and still would be able to use this system to visualize, monitor, and analyze their performance.
{"title":"A Web-based Interactive and Visualized Approach to Simulations of Operating Systems","authors":"H. Pham","doi":"10.1109/ICICT55905.2022.00033","DOIUrl":"https://doi.org/10.1109/ICICT55905.2022.00033","url":null,"abstract":"Computer operating systems are complex systems with many components which are difficult to monitor, analyze, and learn. Operating systems simulators can help to overcome these challenges. This paper provides a brief review of operating system simulators and proposes Web VizOS, a web-based framework for interactive simulations and visualizations of operating system main concepts and mechanisms. This framework can be used for: (i)development, research, and optimization of each individual operating system mechanism; (ii)comparison and analysis of alternative methods in the same category; and (iii) study and research of operating systems in whole. This framework is designed to be comprehensive and can include all fundamental mechanisms in process management, memory management, I/O control, file and disk management. This simulation framework is open and flexible enough that users can develop their own codes for operating system mechanisms independently and still would be able to use this system to visualize, monitor, and analyze their performance.","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121481465","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}
Pub Date : 2022-03-01DOI: 10.1109/ICICT55905.2022.00031
Yamuna Prasad, Nitin
This paper presents an improved COVID19 prediction model using chest X-Ray images with evolutionary algorithm based ensemble learning. The proposed model uses the transfer learning approach with state-of-the-art pre-trained models for training in isolation. Following the fine-tuning of the models, ensemble of the models is used for inferencing. The weight of the ensemble models are learned by the Differential Evolutional (DE) algorithm. The proposed model exploits the importance of each model in COVID19 inferencing. The proposed model is experimented on COVIDx-CXR2 dataset. Our study shows that the proposed EnsembleNet model outperforms the individual state-of-the-art models in terms of generalization accuracy.
{"title":"EnsembleNet: An improved COVID19 Prediction Model using Chest X-Ray Images","authors":"Yamuna Prasad, Nitin","doi":"10.1109/ICICT55905.2022.00031","DOIUrl":"https://doi.org/10.1109/ICICT55905.2022.00031","url":null,"abstract":"This paper presents an improved COVID19 prediction model using chest X-Ray images with evolutionary algorithm based ensemble learning. The proposed model uses the transfer learning approach with state-of-the-art pre-trained models for training in isolation. Following the fine-tuning of the models, ensemble of the models is used for inferencing. The weight of the ensemble models are learned by the Differential Evolutional (DE) algorithm. The proposed model exploits the importance of each model in COVID19 inferencing. The proposed model is experimented on COVIDx-CXR2 dataset. Our study shows that the proposed EnsembleNet model outperforms the individual state-of-the-art models in terms of generalization accuracy.","PeriodicalId":273927,"journal":{"name":"2022 5th International Conference on Information and Computer Technologies (ICICT)","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116595375","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：