The creation of cultural iconography that may reflect national culture and encourage individuals to identify with Chinese culture has always been a difficult issue. In this study, we present a symbolic creation framework for Chinese national cultural identity constructed from visual pictures using generative adversarial networks (GAN). To enhance the structure collapse phenomena of generative adversarial systems, form search regular procedure and generator cross-loss factors on the basis of GAN should be combined. To enhance the real-time efficiency of the model by lowering the parameters in the model, the conventional convolutional component of the generator in the system’s architecture is substituted with a significant recoverable convolution. The notions of iconography and character as they relate to symbols are discussed in this essay. It also advises using iconography as a technique of symbolic imagery to give emergent symbols identity. The design in this study may create significant performance ethnic cultural symbols while preserving superior temporal performance, according to the findings of rigorous testing on real datasets, which may have practical application value. The accuracy, precision, recall, and F1 of the system in this study are 91.54%, 89.02%, 90.96%, and 87.48%.
如何创造能反映民族文化并鼓励个人认同中国文化的文化图标一直是个难题。在本研究中，我们提出了一个利用生成对抗网络（GAN）从视觉图片构建中国民族文化认同的符号创建框架。为了增强生成式对抗系统的结构崩溃现象，应在 GAN 的基础上结合形式搜索规则程序和生成器交叉损失因子。为了通过降低模型中的参数来提高模型的实时效率，系统架构中生成器的传统卷积成分被重要的可恢复卷积所取代。本文讨论了与符号相关的图标和特征概念。 文章还建议使用图标作为符号图像技术，赋予新出现的符号以特性。 根据在真实数据集上进行的严格测试结果，本研究中的设计可能会创造出性能显著的民族文化符号，同时保持卓越的时间性能，这可能具有实际应用价值。本研究中系统的准确度、精确度、召回率和 F1 分别为 91.54%、89.02%、90.96% 和 87.48%。
{"title":"The role of iconography in shaping Chinese national identity: Analyzing its representation in visual media and political propaganda","authors":"HuiXia Zhen, Bo Han","doi":"10.32629/jai.v7i3.1516","DOIUrl":"https://doi.org/10.32629/jai.v7i3.1516","url":null,"abstract":"The creation of cultural iconography that may reflect national culture and encourage individuals to identify with Chinese culture has always been a difficult issue. In this study, we present a symbolic creation framework for Chinese national cultural identity constructed from visual pictures using generative adversarial networks (GAN). To enhance the structure collapse phenomena of generative adversarial systems, form search regular procedure and generator cross-loss factors on the basis of GAN should be combined. To enhance the real-time efficiency of the model by lowering the parameters in the model, the conventional convolutional component of the generator in the system’s architecture is substituted with a significant recoverable convolution. The notions of iconography and character as they relate to symbols are discussed in this essay. It also advises using iconography as a technique of symbolic imagery to give emergent symbols identity. The design in this study may create significant performance ethnic cultural symbols while preserving superior temporal performance, according to the findings of rigorous testing on real datasets, which may have practical application value. The accuracy, precision, recall, and F1 of the system in this study are 91.54%, 89.02%, 90.96%, and 87.48%.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139529275","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}
Yogesh M. Gajmal, Pranav More, Arvind Jagtap, Kiran Kale
Access control is the most vital aspect of cloud data storage security. Traditional techniques for data distribution as well as access control face noteworthy challenges in the arena of research as a result of extensive abuse and privacy data breaches. The blockchain concept provides security by verifying users by multiple encryption technologies. Collaboration in the cloud improves management but compromises privacy. Consequently, we created an efficient access management and data exchange system for a blockchain-based decentralized cloud. On the basis of an ID and password, the data user (DU) submits a registering request to the data owner (DO). The DO data is incorporated into a transactional blockchain by an encoded master key. The data owner (DO) provides data encryption, and encrypted files are still published to the Interplanetary File System (IPFS). The DO generates ciphertext metadata, which is then published to the transactional blockchain utilizing a secure file location and a secure key. The projected access control and data sharing solution performed better in a decentralized blockchain based cloud, as measured by metrics such as a reduced illegitimate user rate of 5%, and a size blockchain of is 100 and 200, respectively.
访问控制是云数据存储安全最重要的方面。传统的数据分发和访问控制技术在研究领域面临着值得注意的挑战，因为存在大量滥用和隐私数据泄露的情况。区块链概念通过多种加密技术验证用户，从而提供了安全性。云中的协作改善了管理，但却损害了隐私。因此，我们为基于区块链的去中心化云创建了一个高效的访问管理和数据交换系统。根据 ID 和密码，数据用户（DU）向数据所有者（DO）提交注册请求。数据所有者（DO）通过编码主密钥将数据纳入交易区块链。数据所有者（DO）提供数据加密，加密文件仍发布到星际文件系统（IPFS）。DO 生成密文元数据，然后利用安全文件位置和安全密钥将其发布到交易区块链上。预计的访问控制和数据共享解决方案在基于去中心化区块链的云中表现更佳，具体指标包括非法用户率降低 5%，区块链大小分别为 100 和 200。
{"title":"Access control and data sharing mechanism in decentralized cloud using blockchain technology","authors":"Yogesh M. Gajmal, Pranav More, Arvind Jagtap, Kiran Kale","doi":"10.32629/jai.v7i3.1332","DOIUrl":"https://doi.org/10.32629/jai.v7i3.1332","url":null,"abstract":"Access control is the most vital aspect of cloud data storage security. Traditional techniques for data distribution as well as access control face noteworthy challenges in the arena of research as a result of extensive abuse and privacy data breaches. The blockchain concept provides security by verifying users by multiple encryption technologies. Collaboration in the cloud improves management but compromises privacy. Consequently, we created an efficient access management and data exchange system for a blockchain-based decentralized cloud. On the basis of an ID and password, the data user (DU) submits a registering request to the data owner (DO). The DO data is incorporated into a transactional blockchain by an encoded master key. The data owner (DO) provides data encryption, and encrypted files are still published to the Interplanetary File System (IPFS). The DO generates ciphertext metadata, which is then published to the transactional blockchain utilizing a secure file location and a secure key. The projected access control and data sharing solution performed better in a decentralized blockchain based cloud, as measured by metrics such as a reduced illegitimate user rate of 5%, and a size blockchain of is 100 and 200, respectively.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"7 10","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139532488","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}
An important analytical tool for tracking, mapping, and quantifying changes in land use and land cover (LULC) across time serves as the use of machine learning techniques. The environment and human activities both have the potential to change how land is used and covered. Classifying LULC types at different spatial scales has been effectively achieved by models like classification and regression trees (CART), support vector machines (SVM), extreme gradient boosting (XGBoost), and random forests (RF). To prepare images from Landsat before sending and analysis for an aspect of our research, we employed the Google Earth Engine. High-resolution imagery from Google Earth images were used to assess each kind of method and field data collection. Utilizing Geographic Information System (GIS) techniques, LULC fluctuations between 2015 and 2020 were assessed. According to our results, XGBoost, SVM, and CART models proved superior by the RF model regarding categorization precision. Considering the data, we collected between 2015 and 2020, from 11.57 hectares (1.74%) in 2015 to 184.19 hectares (27.65%) in 2020, the barren land experienced the greatest variation, that made an immense effect. Utilizing the support of satellite imagery from the Karaivetti Wetland, our work combines novel GIS techniques and machine learning strategies to LULC monitoring. The created land cover maps provide a vital benchmark that will be useful to authorities in formulating policies, managing for sustainability, and keeping track of degradation.
{"title":"Identifying land use land cover dynamics using machine learning method and GIS approach in Karaivetti, Tamil Nadu","authors":"Thylashri Sivasubramaniyan, Rajalakshmi Nagarnaidu Rajaperumal","doi":"10.32629/jai.v7i3.1333","DOIUrl":"https://doi.org/10.32629/jai.v7i3.1333","url":null,"abstract":"An important analytical tool for tracking, mapping, and quantifying changes in land use and land cover (LULC) across time serves as the use of machine learning techniques. The environment and human activities both have the potential to change how land is used and covered. Classifying LULC types at different spatial scales has been effectively achieved by models like classification and regression trees (CART), support vector machines (SVM), extreme gradient boosting (XGBoost), and random forests (RF). To prepare images from Landsat before sending and analysis for an aspect of our research, we employed the Google Earth Engine. High-resolution imagery from Google Earth images were used to assess each kind of method and field data collection. Utilizing Geographic Information System (GIS) techniques, LULC fluctuations between 2015 and 2020 were assessed. According to our results, XGBoost, SVM, and CART models proved superior by the RF model regarding categorization precision. Considering the data, we collected between 2015 and 2020, from 11.57 hectares (1.74%) in 2015 to 184.19 hectares (27.65%) in 2020, the barren land experienced the greatest variation, that made an immense effect. Utilizing the support of satellite imagery from the Karaivetti Wetland, our work combines novel GIS techniques and machine learning strategies to LULC monitoring. The created land cover maps provide a vital benchmark that will be useful to authorities in formulating policies, managing for sustainability, and keeping track of degradation.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"74 16","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139440471","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, with the continuous development of wireless communication and virtual reality technology, multimedia interaction technology has received more and more attention. However, due to the limitations of bandwidth, delay, packet loss, and other problems in wireless networks, multimedia interaction technology also faces many challenges. In this paper, the virtual reality multimedia interaction technology will be studied in combination with a wireless network, in which the virtual environment is first constructed. Then, the interaction is carried out through data transmission and user interaction, and then the existing virtual reality multimedia interaction is optimized by using a data compression algorithm. In order to test the effectiveness and performance of virtual reality multimedia interaction under different network environments and determine the optimal network environment, this paper compares the multimedia interaction effect under three network environments using a wireless network, mobile network, and Bluetooth network as the research object, with download speed, loading speed, image quality, smoothness and real-time as the test variables, and 10 VR software as the constraints. The research results indicated that, under the same other conditions, taking real-time performance as an example, the delay time and feedback speed of wireless networks were between 26 ms–37 ms and 105 KB/s–115 KB/s. The delay time and feedback speed of mobile networks (3G/4G/5G, generation) were between 66 ms–75 ms and 46 KB/s–55 KB/s, while the delay time and feedback speed of Bluetooth networks were between 120 ms–130 ms and 25 KB/s–35 KB/s; this indicated that VR multimedia interaction technology in the wireless network had better performance. Virtual reality multimedia interaction technology based on wireless networks is a promising high technology that can bring users a more realistic, vivid, and intuitive communication and entertainment experience. At the same time, it will expand the application scenarios and promote the development and progress of the technology.
{"title":"Investigation of virtual reality multimedia interaction technology based on wireless network","authors":"Qing Ma","doi":"10.32629/jai.v7i3.1295","DOIUrl":"https://doi.org/10.32629/jai.v7i3.1295","url":null,"abstract":"In recent years, with the continuous development of wireless communication and virtual reality technology, multimedia interaction technology has received more and more attention. However, due to the limitations of bandwidth, delay, packet loss, and other problems in wireless networks, multimedia interaction technology also faces many challenges. In this paper, the virtual reality multimedia interaction technology will be studied in combination with a wireless network, in which the virtual environment is first constructed. Then, the interaction is carried out through data transmission and user interaction, and then the existing virtual reality multimedia interaction is optimized by using a data compression algorithm. In order to test the effectiveness and performance of virtual reality multimedia interaction under different network environments and determine the optimal network environment, this paper compares the multimedia interaction effect under three network environments using a wireless network, mobile network, and Bluetooth network as the research object, with download speed, loading speed, image quality, smoothness and real-time as the test variables, and 10 VR software as the constraints. The research results indicated that, under the same other conditions, taking real-time performance as an example, the delay time and feedback speed of wireless networks were between 26 ms–37 ms and 105 KB/s–115 KB/s. The delay time and feedback speed of mobile networks (3G/4G/5G, generation) were between 66 ms–75 ms and 46 KB/s–55 KB/s, while the delay time and feedback speed of Bluetooth networks were between 120 ms–130 ms and 25 KB/s–35 KB/s; this indicated that VR multimedia interaction technology in the wireless network had better performance. Virtual reality multimedia interaction technology based on wireless networks is a promising high technology that can bring users a more realistic, vivid, and intuitive communication and entertainment experience. At the same time, it will expand the application scenarios and promote the development and progress of the technology.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"17 9","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139439105","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}
Abdullah Alsokkar, M. Otair, Hamza Essam Alfar, A. Nasereddin, Khaled Aldiabat, L. Abualigah
Call centers handle thousands of incoming calls daily, encompassing a diverse array of categories including product inquiries, complaints, and more. Within these conversations, customers articulate their opinions and interests in the products and services offered. Effectively categorizing and analyzing these calls holds immense importance for organizations, offering a window into their strengths, weaknesses, and gauging customer satisfaction and needs. This paper introduces an innovative approach to extract customer sentiments through an advanced sentiment analysis technique. Leveraging two distinct yet synergistic algorithms—Support Vector Machine (SVM) and Neural Networks (NNs)—on the Kaggle machine-learning platform, our method discerns the polarity of each note, classifying them as positive, negative, or neutral. To enhance the quality of our analysis, we employed Natural Language Processing (NLP) and a range of preprocessing tools, including tokenization. The dataset comprises three thousand notes from various telecommunication companies, authored during real call center interactions. These notes form the basis of a specialized corpus, notable for being composed in the Jordanian dialect. Rigorous training and testing procedures were conducted using this corpus. The results are notable: our proposed algorithms displayed strong performance metrics. SVM yielded a commendable accuracy rate of 66%, while NNs excelled, boasting an impressive accuracy rate of 99.21%. These achievements are substantiated by comprehensive confusion matrices. In conclusion, our research provides a novel and robust framework for customer sentiment analysis in call centers, underpinned by the fusion of SVM and NNs. This technique promises valuable insights into customer feedback, facilitating informed decision-making for businesses seeking to enhance their services and products.
{"title":"Sentiment analysis for Arabic call center notes using machine learning techniques","authors":"Abdullah Alsokkar, M. Otair, Hamza Essam Alfar, A. Nasereddin, Khaled Aldiabat, L. Abualigah","doi":"10.32629/jai.v7i3.940","DOIUrl":"https://doi.org/10.32629/jai.v7i3.940","url":null,"abstract":"Call centers handle thousands of incoming calls daily, encompassing a diverse array of categories including product inquiries, complaints, and more. Within these conversations, customers articulate their opinions and interests in the products and services offered. Effectively categorizing and analyzing these calls holds immense importance for organizations, offering a window into their strengths, weaknesses, and gauging customer satisfaction and needs. This paper introduces an innovative approach to extract customer sentiments through an advanced sentiment analysis technique. Leveraging two distinct yet synergistic algorithms—Support Vector Machine (SVM) and Neural Networks (NNs)—on the Kaggle machine-learning platform, our method discerns the polarity of each note, classifying them as positive, negative, or neutral. To enhance the quality of our analysis, we employed Natural Language Processing (NLP) and a range of preprocessing tools, including tokenization. The dataset comprises three thousand notes from various telecommunication companies, authored during real call center interactions. These notes form the basis of a specialized corpus, notable for being composed in the Jordanian dialect. Rigorous training and testing procedures were conducted using this corpus. The results are notable: our proposed algorithms displayed strong performance metrics. SVM yielded a commendable accuracy rate of 66%, while NNs excelled, boasting an impressive accuracy rate of 99.21%. These achievements are substantiated by comprehensive confusion matrices. In conclusion, our research provides a novel and robust framework for customer sentiment analysis in call centers, underpinned by the fusion of SVM and NNs. This technique promises valuable insights into customer feedback, facilitating informed decision-making for businesses seeking to enhance their services and products.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"50 5","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139442039","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}
Sustainability challenges and ICT perspectives are at the heart of current thinking on global economic and social development. The current development and growth process is based on unsustainable foundations due to irresponsible resource consumption and negative environmental impacts as well as greenhouse gas emissions. People need to find ways to integrate the digital economy and the sustainability of the green economy. Therefore, this paper firstly describes the intersection of digital economy and green economy, secondly introduces the security system of digital economy and green economy, then based on this, the SURF (Speeded-up robust features) algorithm is used to locate and improve the data aggregation system of digital economy and green economy, and finally, the algorithm simulation experiment is conducted. The experimental results found that the aggregation algorithm based on digital economy and green economy has 19% higher accuracy than the traditional algorithm. At the same time, the calculation speed is increased by three to five times. The above results show that the SURF algorithm is applied to the sustainable development research of digital economy and green economy with significant effect.
{"title":"Synergy of digital economy and green economy in sustainable development policy","authors":"Qiqi Lin, Ping Zhou","doi":"10.32629/jai.v7i3.1299","DOIUrl":"https://doi.org/10.32629/jai.v7i3.1299","url":null,"abstract":"Sustainability challenges and ICT perspectives are at the heart of current thinking on global economic and social development. The current development and growth process is based on unsustainable foundations due to irresponsible resource consumption and negative environmental impacts as well as greenhouse gas emissions. People need to find ways to integrate the digital economy and the sustainability of the green economy. Therefore, this paper firstly describes the intersection of digital economy and green economy, secondly introduces the security system of digital economy and green economy, then based on this, the SURF (Speeded-up robust features) algorithm is used to locate and improve the data aggregation system of digital economy and green economy, and finally, the algorithm simulation experiment is conducted. The experimental results found that the aggregation algorithm based on digital economy and green economy has 19% higher accuracy than the traditional algorithm. At the same time, the calculation speed is increased by three to five times. The above results show that the SURF algorithm is applied to the sustainable development research of digital economy and green economy with significant effect.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"22 5","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139443233","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}
Osamah Ibrahim Khalaf, D. Anand, G. Abdulsahib, G. R. Chandra
Protecting the mobile cloud computing system from the cyber-threats is the most crucial and demanding problems in recent days. Due to the rapid growth of internet technology, it is more essential to ensure secure the mobile cloud systems against the network intrusions. In the existing works, various intrusion detection system (IDS) frameworks have been developed for mobile cloud security, which are mainly focusing on utilizing the optimization and classification algorithms for designing the security frameworks. Still, some of the challenges associated to the existing works are complex to understand the system model, educed convergence rate, inability to handle complex datasets, and high time cost. Therefore, this research work motivates to design and develop a computationally efficient IDS framework for improving the mobile cloud systems security. Here, an intrinsic collateral normalization (InCoN) algorithm is implemented at first for generating the quality improved datasets. Consequently, the coherent salp swarm optimization (CSSO) technique is deployed for selecting the most relevant features used for intrusion prediction and categorization. Finally, the deep reinforced neural network (DRNN) mechanism is implemented for accurately detecting the type of intrusion by properly training and testing the optimal features. During validation, the findings of the CSSO-DRNN technique are assessed and verified by utilizing various QoS parameters.
{"title":"A coherent salp swarm optimization based deep reinforced neuralnet work algorithm for securing the mobile cloud systems","authors":"Osamah Ibrahim Khalaf, D. Anand, G. Abdulsahib, G. R. Chandra","doi":"10.32629/jai.v7i3.654","DOIUrl":"https://doi.org/10.32629/jai.v7i3.654","url":null,"abstract":"Protecting the mobile cloud computing system from the cyber-threats is the most crucial and demanding problems in recent days. Due to the rapid growth of internet technology, it is more essential to ensure secure the mobile cloud systems against the network intrusions. In the existing works, various intrusion detection system (IDS) frameworks have been developed for mobile cloud security, which are mainly focusing on utilizing the optimization and classification algorithms for designing the security frameworks. Still, some of the challenges associated to the existing works are complex to understand the system model, educed convergence rate, inability to handle complex datasets, and high time cost. Therefore, this research work motivates to design and develop a computationally efficient IDS framework for improving the mobile cloud systems security. Here, an intrinsic collateral normalization (InCoN) algorithm is implemented at first for generating the quality improved datasets. Consequently, the coherent salp swarm optimization (CSSO) technique is deployed for selecting the most relevant features used for intrusion prediction and categorization. Finally, the deep reinforced neural network (DRNN) mechanism is implemented for accurately detecting the type of intrusion by properly training and testing the optimal features. During validation, the findings of the CSSO-DRNN technique are assessed and verified by utilizing various QoS parameters.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"12 21","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139443377","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}
Providing meaningful classification for each pixel in an image is a primary goal of computer vision, and the tasks of object classification and semantic segmentation are among the field’s greatest challenges. To improve object classification, this study presents a novel method that combines semantic segmentation with dynamic convolution layer-based optimization techniques. In the proposed method, a Refined Convolution Neural Network (R-CNN) is used, which uses non-extensive entropy to dynamically increase the size of its convolutional layers. The Common Objects in Context (COCO) dataset is used to assess the performance of the model. The model performs exceptionally well at different Intersections over Union (IoU) cutoffs, with average precision values of 40.1, 61.9, and 45.4, respectively, for Average Precision (AP), AP50, and AP75. These results demonstrate the model’s efficiency in discriminating between various image contents. Additionally, the model predicts an image’s outcome on average in just 0.901 s. The model has been proven to be superior through various performance evaluation parameters, showing an average mean precision of 91.78%. This study demonstrates the power of combining dynamic convolution layers with semantic segmentation to improve object classification accuracy, a key component in the development of computer vision applications.
{"title":"Dynamic convolution layer based optimization techniques for object classification and semantic segmentation","authors":"Jaswinder Singh, B. K. Sharma","doi":"10.32629/jai.v7i3.944","DOIUrl":"https://doi.org/10.32629/jai.v7i3.944","url":null,"abstract":"Providing meaningful classification for each pixel in an image is a primary goal of computer vision, and the tasks of object classification and semantic segmentation are among the field’s greatest challenges. To improve object classification, this study presents a novel method that combines semantic segmentation with dynamic convolution layer-based optimization techniques. In the proposed method, a Refined Convolution Neural Network (R-CNN) is used, which uses non-extensive entropy to dynamically increase the size of its convolutional layers. The Common Objects in Context (COCO) dataset is used to assess the performance of the model. The model performs exceptionally well at different Intersections over Union (IoU) cutoffs, with average precision values of 40.1, 61.9, and 45.4, respectively, for Average Precision (AP), AP50, and AP75. These results demonstrate the model’s efficiency in discriminating between various image contents. Additionally, the model predicts an image’s outcome on average in just 0.901 s. The model has been proven to be superior through various performance evaluation parameters, showing an average mean precision of 91.78%. This study demonstrates the power of combining dynamic convolution layers with semantic segmentation to improve object classification accuracy, a key component in the development of computer vision applications.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"25 11","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139443084","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}
The purpose of this study is to develop a disease prediction model that can evaluate diagnostic test results based on a machine learning model and big data analysis algorithms for automated judgment of health chuck-up results. The research method used the catboost algorithm for data pretreatment and analysis. The original data was divided into learning data and test data to ensure 21,140 effective data consisting of 27 properties and to develop and utilize predictive models. Learning data was used as input data for the development of predictive models, and the test data was divided into data for the performance evaluation of the predictive model. Random forest analysis algorithms were used to analyze testing and determination accuracy that affect disease determination, and forecasting model performance analysis was analyzed by accuracy, ROC (ROC) Area, Confusion Matrix, Precision, and Recall indicators. As a result of random forest analysis, both diabetes and two -ventilation diseases were analyzed to be used as a commercial platform model by analyzing more than 90% forecast accuracy. The results of this study found that using big data analysis and machine learning, it is possible to determine and predict specific diseases based on health check-up data.
{"title":"Study on prediction and diagnosis AI model of frequent chronic diseases based on health checkup big data","authors":"Jae Young Park, Jai-Woo Oh","doi":"10.32629/jai.v7i3.999","DOIUrl":"https://doi.org/10.32629/jai.v7i3.999","url":null,"abstract":"The purpose of this study is to develop a disease prediction model that can evaluate diagnostic test results based on a machine learning model and big data analysis algorithms for automated judgment of health chuck-up results. The research method used the catboost algorithm for data pretreatment and analysis. The original data was divided into learning data and test data to ensure 21,140 effective data consisting of 27 properties and to develop and utilize predictive models. Learning data was used as input data for the development of predictive models, and the test data was divided into data for the performance evaluation of the predictive model. Random forest analysis algorithms were used to analyze testing and determination accuracy that affect disease determination, and forecasting model performance analysis was analyzed by accuracy, ROC (ROC) Area, Confusion Matrix, Precision, and Recall indicators. As a result of random forest analysis, both diabetes and two -ventilation diseases were analyzed to be used as a commercial platform model by analyzing more than 90% forecast accuracy. The results of this study found that using big data analysis and machine learning, it is possible to determine and predict specific diseases based on health check-up data.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"15 9","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139445636","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}
Amid a lot of evolutionary methods (EMs), differential evolution (DE) is broadly used for various optimization issues. Though, it has rare shortcomings such as slow convergence, stagnation etc. Likewise, mutation and its control factor choice for DE is extremely inspiring for enhanced optimization. To increase the exploration competence of DE, a modified-DE (M-DE) is advised in this paper. It implemented a new mutation system, thru the perception of particle swarm optimization, to further trade off the population diversity. Meanwhile, centered on time-varying structure, new mutant control parameters incorporated with the suggested mutation scheme, to escaping local optima and keep evolving. Using the features of memory and robustly altered control parameters, exploitation and exploration ability of M-DE is well-adjusted. Also, admitted features of M-DE algorithm follows to speeding up convergence significantly. Finally, to verify the effectiveness of M-DE, groups of assessments have been piloted on six unimodal and seven multimodal benchmark suites. Performance of M-DE compared with different peer DE algorithms. According the investigational results, efficiency of the suggested M-DE technique has been confirmed.
在众多进化方法（EM）中，微分进化法（DE）被广泛应用于各种优化问题。尽管如此，它也存在收敛速度慢、停滞不前等缺点。同样，微分进化论的突变及其控制因子的选择对增强优化也极具启发性。为了提高 DE 的探索能力，本文提出了一种修正 DE（M-DE）。它通过对粒子群优化的感知，实施了一种新的突变系统，以进一步权衡种群的多样性。同时，以时变结构为中心，将新的突变控制参数与建议的突变方案相结合，以摆脱局部最优状态并不断进化。利用记忆和稳健改变控制参数的特点，M-DE 的开发和探索能力得到了很好的调整。此外，M-DE 算法还具有收敛速度快的特点。最后，为了验证 M-DE 的有效性，我们在六个单模态和七个多模态基准套件上进行了试验性评估。将 M-DE 的性能与不同的同行 DE 算法进行比较。根据调查结果，建议的 M-DE 技术的效率得到了证实。
{"title":"Enhancing differential evolution through a modified mutation strategy for unimodal and multimodal problem optimization","authors":"Pooja Tiwari, Vishnu Narayan Mishra, Raghav Prasad Parouha","doi":"10.32629/jai.v7i3.1103","DOIUrl":"https://doi.org/10.32629/jai.v7i3.1103","url":null,"abstract":"Amid a lot of evolutionary methods (EMs), differential evolution (DE) is broadly used for various optimization issues. Though, it has rare shortcomings such as slow convergence, stagnation etc. Likewise, mutation and its control factor choice for DE is extremely inspiring for enhanced optimization. To increase the exploration competence of DE, a modified-DE (M-DE) is advised in this paper. It implemented a new mutation system, thru the perception of particle swarm optimization, to further trade off the population diversity. Meanwhile, centered on time-varying structure, new mutant control parameters incorporated with the suggested mutation scheme, to escaping local optima and keep evolving. Using the features of memory and robustly altered control parameters, exploitation and exploration ability of M-DE is well-adjusted. Also, admitted features of M-DE algorithm follows to speeding up convergence significantly. Finally, to verify the effectiveness of M-DE, groups of assessments have been piloted on six unimodal and seven multimodal benchmark suites. Performance of M-DE compared with different peer DE algorithms. According the investigational results, efficiency of the suggested M-DE technique has been confirmed.","PeriodicalId":508223,"journal":{"name":"Journal of Autonomous Intelligence","volume":"57 47","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139447148","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：