Pub Date : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194902
Hua Qu, Jiaqi Li, Yanpeng Zhang
This paper proposes a long short-term memory network (FTS-LSTM) prediction model based on fuzzy time series to improve the prediction accuracy of time series. First, the fuzzy C-means clustering FCM algorithm is used to classify the time series to form a fuzzy time series and obtain the membership matrix. Second, the LSTM net-work prediction model is constructed, and the FTS-LSTM network prediction model is proposed. The previously obtained membership is used as the full connection. The weight of the layer and its membership as the weight remain unchanged. This FTS-LSTM network prediction model not only considers the non-linearity and non-stationarity of the time series, but also resolves the inherent uncertainty and ambiguity of the data. Simulation results show that the FTS-LSTM network-based prediction model has faster training speed, higher prediction accuracy, and better prediction effect on time series with large ambiguities.
{"title":"Long Short-term Memory Network Prediction Model Based on Fuzzy Time Series","authors":"Hua Qu, Jiaqi Li, Yanpeng Zhang","doi":"10.1109/ICAIIS49377.2020.9194902","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194902","url":null,"abstract":"This paper proposes a long short-term memory network (FTS-LSTM) prediction model based on fuzzy time series to improve the prediction accuracy of time series. First, the fuzzy C-means clustering FCM algorithm is used to classify the time series to form a fuzzy time series and obtain the membership matrix. Second, the LSTM net-work prediction model is constructed, and the FTS-LSTM network prediction model is proposed. The previously obtained membership is used as the full connection. The weight of the layer and its membership as the weight remain unchanged. This FTS-LSTM network prediction model not only considers the non-linearity and non-stationarity of the time series, but also resolves the inherent uncertainty and ambiguity of the data. Simulation results show that the FTS-LSTM network-based prediction model has faster training speed, higher prediction accuracy, and better prediction effect on time series with large ambiguities.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131944882","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 : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194929
Dongxu Quan, Xiameng Lian
Phase interpolation based digital clock data recovery are widely adopted in Serdes design because of capability of dealing with burst mode. In this paper, a two-stage phase interpolator utilizing IQ clock are proposed. The tail current in first stage can be trimmed to equalized the amplitude difference caused by first stage interpolation. The second stage operates 8-step phase interpolation by using clock with 45° difference. The Circuit is implemented in HLMC 55nmdr process. The DNL is 0.8LSB, INL is 2LSB, typical power consumption is 36.36mW@1.2V. PI operating frequency is 2.5G and its control logic operates at 312.5MHz.
{"title":"A 2-Stage Phase Interpolator Used in Clock Data Recovery Circuit","authors":"Dongxu Quan, Xiameng Lian","doi":"10.1109/ICAIIS49377.2020.9194929","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194929","url":null,"abstract":"Phase interpolation based digital clock data recovery are widely adopted in Serdes design because of capability of dealing with burst mode. In this paper, a two-stage phase interpolator utilizing IQ clock are proposed. The tail current in first stage can be trimmed to equalized the amplitude difference caused by first stage interpolation. The second stage operates 8-step phase interpolation by using clock with 45° difference. The Circuit is implemented in HLMC 55nmdr process. The DNL is 0.8LSB, INL is 2LSB, typical power consumption is 36.36mW@1.2V. PI operating frequency is 2.5G and its control logic operates at 312.5MHz.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132423265","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 : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194900
Darui He, Jianlong Sun, Yan Li, Fangyuan Tian, Yiran Chen, Guodao Tong, Xisong Chen, Qipeng Shen, Zhibo Lian
This paper studies a thermal runaway warning system for the safety management system of lithium iron phosphate battery for energy storage. The entire process of thermal runaway is analyzed and controlled according to the process, including temperature warnings, gas warnings, smoke and infrared warnings. Then, the problem of position and threshold setting of the warning sensors are studied. Finally, a hierarchical warning system is established and a communication architecture diagram of system warning is constructed. It is shown that the system can quickly locate the area where the battery pack is out of control, and quickly perform corresponding disconnection, firefighting and alarm operations to ensure the safe and stable operation of the battery storage power station.
{"title":"Thermal Runaway Warning Based on Safety Management System of Lithium Iron Phosphate Battery for Energy Storage","authors":"Darui He, Jianlong Sun, Yan Li, Fangyuan Tian, Yiran Chen, Guodao Tong, Xisong Chen, Qipeng Shen, Zhibo Lian","doi":"10.1109/ICAIIS49377.2020.9194900","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194900","url":null,"abstract":"This paper studies a thermal runaway warning system for the safety management system of lithium iron phosphate battery for energy storage. The entire process of thermal runaway is analyzed and controlled according to the process, including temperature warnings, gas warnings, smoke and infrared warnings. Then, the problem of position and threshold setting of the warning sensors are studied. Finally, a hierarchical warning system is established and a communication architecture diagram of system warning is constructed. It is shown that the system can quickly locate the area where the battery pack is out of control, and quickly perform corresponding disconnection, firefighting and alarm operations to ensure the safe and stable operation of the battery storage power station.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"129 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130054895","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 : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194856
Chenquan Huang, Weihang Wu, Zhihua Lei
We present an efficient, flexible, fast and accurate framework for real-time instance segmentation. We call it Efficient Instance Segmentation Network, denoted as EISNET. Our method is motivated by the Mask R-CNN and YOLACT. Mask R-CNN enables instance segmentation by adding an extra branch at the Faster R-CNN framework to produce mask for each object. Due to limitation of the inefficiency of two stage detector, Mask R-CNN is not suitable for real-time scene. We therefore propose EISNET which enables instance segmentation by adding two branches to the one-stage detector-RetinaNet. We call it Efficient since we use modified EfficientNet as the backbone of our framework, which results in high accuracy with few parameters and FLOPS. In addition, we provide a modified bi-directional FPN (Feature Pyramid Network) module, which thus allows efficient multi-scale feature fusion. Given the credit to these design techniques, our EISNET achieves 31.2 mAP with only 17.2M parameters and 3.5B FLOPS on the COCO dataset. More significantly, our model can achieve more than 35 FPS on single 1080Ti GPU, which fits the most real-time requirements. With a better GPU, we could even achieve higher mAP while keeping the real-time property of more than 30 FPS.
{"title":"Efficient Instance Segmentation Network","authors":"Chenquan Huang, Weihang Wu, Zhihua Lei","doi":"10.1109/ICAIIS49377.2020.9194856","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194856","url":null,"abstract":"We present an efficient, flexible, fast and accurate framework for real-time instance segmentation. We call it Efficient Instance Segmentation Network, denoted as EISNET. Our method is motivated by the Mask R-CNN and YOLACT. Mask R-CNN enables instance segmentation by adding an extra branch at the Faster R-CNN framework to produce mask for each object. Due to limitation of the inefficiency of two stage detector, Mask R-CNN is not suitable for real-time scene. We therefore propose EISNET which enables instance segmentation by adding two branches to the one-stage detector-RetinaNet. We call it Efficient since we use modified EfficientNet as the backbone of our framework, which results in high accuracy with few parameters and FLOPS. In addition, we provide a modified bi-directional FPN (Feature Pyramid Network) module, which thus allows efficient multi-scale feature fusion. Given the credit to these design techniques, our EISNET achieves 31.2 mAP with only 17.2M parameters and 3.5B FLOPS on the COCO dataset. More significantly, our model can achieve more than 35 FPS on single 1080Ti GPU, which fits the most real-time requirements. With a better GPU, we could even achieve higher mAP while keeping the real-time property of more than 30 FPS.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134112451","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 : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194896
Gengyi Xiao
Internet big data is related to our personal privacy and property security. In serious cases, it also affects the security of confidential information of related companies. Therefore, big data must be protected to prevent criminals from stealing our personal privacy, property information and corporate secrets. Based on this research background, the paper designs the design of computer network security defines system. After the system design is implemented, the system is tested accordingly. According to the test results, the computer network security defines system designed in this paper can actively detect and effectively prevent security threats in the network, thereby ensuring that the network can Normal and safe operation. The computing network security defines system can also provide effective ideas for future network security protection and achieve further expansion of security defines.
{"title":"Research on Computer Network Information Security Based on Big Data Technology","authors":"Gengyi Xiao","doi":"10.1109/ICAIIS49377.2020.9194896","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194896","url":null,"abstract":"Internet big data is related to our personal privacy and property security. In serious cases, it also affects the security of confidential information of related companies. Therefore, big data must be protected to prevent criminals from stealing our personal privacy, property information and corporate secrets. Based on this research background, the paper designs the design of computer network security defines system. After the system design is implemented, the system is tested accordingly. According to the test results, the computer network security defines system designed in this paper can actively detect and effectively prevent security threats in the network, thereby ensuring that the network can Normal and safe operation. The computing network security defines system can also provide effective ideas for future network security protection and achieve further expansion of security defines.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131559582","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 : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194868
Donglong Wang, Jincheng Zhao
For improving the stability of the engine driven self-excitation cage asynchronous generation system (EDS-CAGS) on the condition of the large range speed variation and the impact load, reducing the pulsation of the direct torque control (DTC), a new voltage space vector DTC is proposed based on the Super-Twisting control of the voltage-linkage outer subsystem and the K-class direct feedback linearization control of the current inner subsystem. EDS-CAGS simulation results show that, on the condition of the large range speed variation and the impact load, compared with the traditional voltage outer-loop and current inner-loop DTC method, by the new Super-Twisting sliding mode control method, the overshoot of the DC output voltage is reduced significantly, the response speed of the torque is fastened, the derivative values of the sliding mode variables can be convergence, and the robust stability of the EDS-CAGS is enhanced.
{"title":"Self-excitation Control of Squirrel-Cage Induction Motor based on Super-Twisting Sliding Mode Algorithm","authors":"Donglong Wang, Jincheng Zhao","doi":"10.1109/ICAIIS49377.2020.9194868","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194868","url":null,"abstract":"For improving the stability of the engine driven self-excitation cage asynchronous generation system (EDS-CAGS) on the condition of the large range speed variation and the impact load, reducing the pulsation of the direct torque control (DTC), a new voltage space vector DTC is proposed based on the Super-Twisting control of the voltage-linkage outer subsystem and the K-class direct feedback linearization control of the current inner subsystem. EDS-CAGS simulation results show that, on the condition of the large range speed variation and the impact load, compared with the traditional voltage outer-loop and current inner-loop DTC method, by the new Super-Twisting sliding mode control method, the overshoot of the DC output voltage is reduced significantly, the response speed of the torque is fastened, the derivative values of the sliding mode variables can be convergence, and the robust stability of the EDS-CAGS is enhanced.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131802810","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 : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194807
Zeqian Liu, Yiguo Ji, Lin Yang, Chunyan Tian
Aiming at meeting the operational need of guidance improvement for conventional aerial bomb, the overall structure of pitch channel control system of guidance improvement for conventional aerial bomb is studied, and the selection method of the performance index of the pitch channel stability control loop is determined, and the design requirements are analyzed. A stable loop control model for pitch channel based on pseudo attacking-angle feedback three loops method is established. The performance of the stability control loop is simulated by means of parameter freezing method. The simulation results show that the design of the pitch channel control system achieves the performance requirements.
{"title":"Design and Research on Guidance Improved Conventional Aerial Bomb Pitch Channel Control System","authors":"Zeqian Liu, Yiguo Ji, Lin Yang, Chunyan Tian","doi":"10.1109/ICAIIS49377.2020.9194807","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194807","url":null,"abstract":"Aiming at meeting the operational need of guidance improvement for conventional aerial bomb, the overall structure of pitch channel control system of guidance improvement for conventional aerial bomb is studied, and the selection method of the performance index of the pitch channel stability control loop is determined, and the design requirements are analyzed. A stable loop control model for pitch channel based on pseudo attacking-angle feedback three loops method is established. The performance of the stability control loop is simulated by means of parameter freezing method. The simulation results show that the design of the pitch channel control system achieves the performance requirements.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133410631","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 : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194849
Junhui Liu, Yajuan Jia, Yaya Wang, Juanjuan Wang
In order to reduce the energy consumption of vehicles in urban road systems, it is necessary to improve the traffic capacity of urban intersections from the perspective of reducing energy consumption signalized intersections that proposed the use of an improved genetic algorithm, and it is to optimize the intersection with time-based approach. More, it is a goal to study the cumulative energy consumption and a simulation experiment, while the simulation results show that this method can effectively reduce the accumulated energy consumption and obtain better traffic signal control effect.
{"title":"Cumulative Energy Consumption Analysis of Signal Intersections Based on Improved Genetic Algorithm","authors":"Junhui Liu, Yajuan Jia, Yaya Wang, Juanjuan Wang","doi":"10.1109/ICAIIS49377.2020.9194849","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194849","url":null,"abstract":"In order to reduce the energy consumption of vehicles in urban road systems, it is necessary to improve the traffic capacity of urban intersections from the perspective of reducing energy consumption signalized intersections that proposed the use of an improved genetic algorithm, and it is to optimize the intersection with time-based approach. More, it is a goal to study the cumulative energy consumption and a simulation experiment, while the simulation results show that this method can effectively reduce the accumulated energy consumption and obtain better traffic signal control effect.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"5 4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133211367","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 : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194904
Xiaomin Zhao
In order to improve the teaching evaluation expert system intelligent and real-time performance, reduce the teaching expert evaluation system output error rate, improve the accuracy of teaching expert evaluation, a teaching expert evaluation system design method is proposed based on Internet and artificial intelligence. The overall design framework for teaching expert evaluation system, the design of network teaching the expert evaluation system of artificial intelligence chaos control method using a one-way chain network, the transmission control protocol for online teaching evaluation and monitoring information recognition, to improve the real-time transmission capability assessment information teaching, to collect multimedia monitoring information into the signal conditioning module, multimedia information monitoring and scheduling in the open the application programming interface. Combined with the artificial intelligent control method to realize remote control teaching expert evaluation system, through the PIC bus will be teaching expert evaluation data acquisition to the PC machine, DSP after receiving the signal after signal processing and playback, the artificial intelligence control of teaching expert evaluation system is realized. The test results show that this method of teaching design expert system with artificial intelligence, it has good data information fusion ability.
{"title":"Design of Teaching Expert Evaluation System Based on Artificial Intelligence","authors":"Xiaomin Zhao","doi":"10.1109/ICAIIS49377.2020.9194904","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194904","url":null,"abstract":"In order to improve the teaching evaluation expert system intelligent and real-time performance, reduce the teaching expert evaluation system output error rate, improve the accuracy of teaching expert evaluation, a teaching expert evaluation system design method is proposed based on Internet and artificial intelligence. The overall design framework for teaching expert evaluation system, the design of network teaching the expert evaluation system of artificial intelligence chaos control method using a one-way chain network, the transmission control protocol for online teaching evaluation and monitoring information recognition, to improve the real-time transmission capability assessment information teaching, to collect multimedia monitoring information into the signal conditioning module, multimedia information monitoring and scheduling in the open the application programming interface. Combined with the artificial intelligent control method to realize remote control teaching expert evaluation system, through the PIC bus will be teaching expert evaluation data acquisition to the PC machine, DSP after receiving the signal after signal processing and playback, the artificial intelligence control of teaching expert evaluation system is realized. The test results show that this method of teaching design expert system with artificial intelligence, it has good data information fusion ability.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"38 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127800552","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 : 2020-03-01DOI: 10.1109/ICAIIS49377.2020.9194812
Yongjun Zhang, Siyu Yan
This paper proposes an improved Cost-Effective Active Learning (CEAL) method for Deep Image Classification: Tri-CEAL, which was based on the Tri-training algorithm. By implementing the semi-supervised learning Tri-Training algorithm in CEAL, Tri-CEAL can use semi-supervised classification to select high-confidence samples in unlabeled samples for feature learning. At the same time, the active learning strategy in CEAL was improved to an active learning algorithm based on voting entropy, in which unlabeled samples with high information value are selected for manual labeling based on voting entropy. The classification experiments of Tri-CEAL algorithm and CEAL algorithm on CIFAR-10 indicate that the Tri-CEAL significantly reduces the workload of manually labeling samples and has better generalization performance on image classification problems.
{"title":"Semi-supervised active learning image classification method based on Tri-Training algorithm","authors":"Yongjun Zhang, Siyu Yan","doi":"10.1109/ICAIIS49377.2020.9194812","DOIUrl":"https://doi.org/10.1109/ICAIIS49377.2020.9194812","url":null,"abstract":"This paper proposes an improved Cost-Effective Active Learning (CEAL) method for Deep Image Classification: Tri-CEAL, which was based on the Tri-training algorithm. By implementing the semi-supervised learning Tri-Training algorithm in CEAL, Tri-CEAL can use semi-supervised classification to select high-confidence samples in unlabeled samples for feature learning. At the same time, the active learning strategy in CEAL was improved to an active learning algorithm based on voting entropy, in which unlabeled samples with high information value are selected for manual labeling based on voting entropy. The classification experiments of Tri-CEAL algorithm and CEAL algorithm on CIFAR-10 indicate that the Tri-CEAL significantly reduces the workload of manually labeling samples and has better generalization performance on image classification problems.","PeriodicalId":416002,"journal":{"name":"2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115511859","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：