Pub Date : 2022-10-14DOI: 10.1109/ICIST55546.2022.9926866
Mengxue Yan, Yan Zhao, Ming Guo, Haoyu Sun, Jianlong Qiu, Feng Zhao
Gait has been shown to be a profound movement in human activities, and gait recognition is a commonly used biometric recognition in recent years. Gait recognition based on wearable sensors has been involved in various application areas. Especially in the area of medical, gait research is an essential issue. The purpose of this paper is to provide a multimodal public dataset for use with gait recognition. The dataset is derived of data from wearable inertial sensors and ECG sensor. Both sensors provide easy-to-operate and low-cost data recording for gait recognition. The gait dataset is based on the data from 15 healthy adults whose lower limbs have neither been injured nor operated on in the past year. Unlike other well-known datasets in the literature, this dataset contains inertial data (built-in gyroscope, accelerometer, geomagnetic field sensor) recorded from the ankle, as well as ECG data from a cardiac sensor. In this paper, the 15 volunteers were asked to walk at their most comfortable pace in four different terrains and complete the test. These four kinds of terrains are: flat land, sand, grassland and blind road. In addition, in order to verify the effectiveness of this multimodal dataset, this paper uses deep learning to identify the gait patterns of four terrains, and the recognition rate reaches 82%.
{"title":"A Multimodal Dataset for Gait Recognition in Different Terrains using Wearable Sensors","authors":"Mengxue Yan, Yan Zhao, Ming Guo, Haoyu Sun, Jianlong Qiu, Feng Zhao","doi":"10.1109/ICIST55546.2022.9926866","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926866","url":null,"abstract":"Gait has been shown to be a profound movement in human activities, and gait recognition is a commonly used biometric recognition in recent years. Gait recognition based on wearable sensors has been involved in various application areas. Especially in the area of medical, gait research is an essential issue. The purpose of this paper is to provide a multimodal public dataset for use with gait recognition. The dataset is derived of data from wearable inertial sensors and ECG sensor. Both sensors provide easy-to-operate and low-cost data recording for gait recognition. The gait dataset is based on the data from 15 healthy adults whose lower limbs have neither been injured nor operated on in the past year. Unlike other well-known datasets in the literature, this dataset contains inertial data (built-in gyroscope, accelerometer, geomagnetic field sensor) recorded from the ankle, as well as ECG data from a cardiac sensor. In this paper, the 15 volunteers were asked to walk at their most comfortable pace in four different terrains and complete the test. These four kinds of terrains are: flat land, sand, grassland and blind road. In addition, in order to verify the effectiveness of this multimodal dataset, this paper uses deep learning to identify the gait patterns of four terrains, and the recognition rate reaches 82%.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132300587","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, emotion recognition technology has been widely used in emotion change perception and mental illness diagnosis. Previous methods are mainly based on single-task learning strategies, which are unable to fuse multimodal features and remove redundant information. This paper proposes an emotion recognition model ER-MRL, which is based on multimodal representation learning. ER-MRL vectorizes the multimodal emotion data through encoders based on neural networks. The gate mechanism is used for multimodal feature selection. On this basis, ER-MRL calculates the modality specific and modality invariant representation for each emotion category. The Transformer model and multihead self-attention layer are applied to multimodal feature fusion. ER-MRL figures out the prediction result through the tower layer based on fully connected neural networks. Experimental results on the CMU-MOSI dataset show that ER-MRL has better performance on emotion recognition than previous methods.
{"title":"ER-MRL: Emotion Recognition based on Multimodal Representation Learning","authors":"Xiaoding Guo, Yadi Wang, Zhijun Miao, Xiaojin Yang, Jinkai Guo, Xianhong Hou, Feifei Zao","doi":"10.1109/ICIST55546.2022.9926848","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926848","url":null,"abstract":"In recent years, emotion recognition technology has been widely used in emotion change perception and mental illness diagnosis. Previous methods are mainly based on single-task learning strategies, which are unable to fuse multimodal features and remove redundant information. This paper proposes an emotion recognition model ER-MRL, which is based on multimodal representation learning. ER-MRL vectorizes the multimodal emotion data through encoders based on neural networks. The gate mechanism is used for multimodal feature selection. On this basis, ER-MRL calculates the modality specific and modality invariant representation for each emotion category. The Transformer model and multihead self-attention layer are applied to multimodal feature fusion. ER-MRL figures out the prediction result through the tower layer based on fully connected neural networks. Experimental results on the CMU-MOSI dataset show that ER-MRL has better performance on emotion recognition than previous methods.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124113589","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 : 2022-10-14DOI: 10.1109/ICIST55546.2022.9926794
Zhantao Liang, Mingming Ha, Derong Liu
In this paper, the value-iteration-based Q-Iearning algorithm with approximation errors is analyzed theoretically. First, based on an upper bound of the approximation errors caused by the Q-function approximator, we get the lower and upper bound functions of the iterative Q-function, which proves that the limit of the approximate Q-function sequence is bounded. Then, we develop a stability condition for the termination of the iterative algorithm, for ensuring that the current control policy derived from the resulting approximate Q-function is stabilizing. Also, we establish an upper bound function of the approximation errors, which is caused by the policy function approximator, to guarantee that the approximate control policy is stabilizing. Finally, the numerical results verifies the theoretical results with a simulation example.
{"title":"Theoretical Analysis of Value-Iteration-Based Q-Learning with Approximation Errors","authors":"Zhantao Liang, Mingming Ha, Derong Liu","doi":"10.1109/ICIST55546.2022.9926794","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926794","url":null,"abstract":"In this paper, the value-iteration-based Q-Iearning algorithm with approximation errors is analyzed theoretically. First, based on an upper bound of the approximation errors caused by the Q-function approximator, we get the lower and upper bound functions of the iterative Q-function, which proves that the limit of the approximate Q-function sequence is bounded. Then, we develop a stability condition for the termination of the iterative algorithm, for ensuring that the current control policy derived from the resulting approximate Q-function is stabilizing. Also, we establish an upper bound function of the approximation errors, which is caused by the policy function approximator, to guarantee that the approximate control policy is stabilizing. Finally, the numerical results verifies the theoretical results with a simulation example.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127275835","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 : 2022-10-14DOI: 10.1109/ICIST55546.2022.9926778
Yijun Zhao, Shaozhi Li, Mian Wang, Xiang Wan, Kun Xia
Land quality is evaluated according to the chemical composition content in soil. The geochemical evaluation of land quality can help users to determine how to use the land, e.g., dynamically manage land resources and adjust the pattern of farming. However, some chemical composition contents are missing in practice. It is necessary to predict the missing chemical composition content for the geochemical evaluation. This paper proposes an adaptive k-nearest-neighbor approach for predicting the chemical composition content in soil. The approach can adaptively determine the similarity between soil samples based on the characteristics of geological background, soil type, land use type and geographical position. According to the similarity, the proposed approach selects the k nearest neighbors of a sample and predicts the missing chemical composition content. The experimental results show that the proposed approach has better accuracy and stability than its competitors.
{"title":"An Adaptive K-Nearest-Neighbor Approach for Predicting Chemical Composition Content in Soil","authors":"Yijun Zhao, Shaozhi Li, Mian Wang, Xiang Wan, Kun Xia","doi":"10.1109/ICIST55546.2022.9926778","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926778","url":null,"abstract":"Land quality is evaluated according to the chemical composition content in soil. The geochemical evaluation of land quality can help users to determine how to use the land, e.g., dynamically manage land resources and adjust the pattern of farming. However, some chemical composition contents are missing in practice. It is necessary to predict the missing chemical composition content for the geochemical evaluation. This paper proposes an adaptive k-nearest-neighbor approach for predicting the chemical composition content in soil. The approach can adaptively determine the similarity between soil samples based on the characteristics of geological background, soil type, land use type and geographical position. According to the similarity, the proposed approach selects the k nearest neighbors of a sample and predicts the missing chemical composition content. The experimental results show that the proposed approach has better accuracy and stability than its competitors.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129172566","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 : 2022-10-14DOI: 10.1109/ICIST55546.2022.9926790
Yiqing Zhang, Wei Zheng, J. Xue, Jianyong Sun
Recovering the transmitted signals in a multiple-input multiple-output (MIMO) system is known to be non-deterministic polynomial hard. It is extremely challenging to obtain a high-quality solution with fairly low computational complexity. To solve the MIMO detection problem effectively, this paper proposes to model it as a time series prediction problem, and a bidirectional temporal convolutional network (Bi- TCN) is designed to address it. In Bi- TCN, the encoder extracts the features of the received signal and the channel matrix by applying non-causal dilated convolution, and the decoder outputs the probability distribution of the recovered transmitted signal in parallel. In the experiments, we compare it with traditional and deep learning-based detectors on both i.i.d. and correlated Rayleigh fading channels, respectively. Experimental results empirically demonstrate that Bi- TCN can achieve near-optimal bit-error-rate (BER) performance with considerably low space complexity.
{"title":"Deep Temporal Sequence Prediction Neural Network for MIMO Detection","authors":"Yiqing Zhang, Wei Zheng, J. Xue, Jianyong Sun","doi":"10.1109/ICIST55546.2022.9926790","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926790","url":null,"abstract":"Recovering the transmitted signals in a multiple-input multiple-output (MIMO) system is known to be non-deterministic polynomial hard. It is extremely challenging to obtain a high-quality solution with fairly low computational complexity. To solve the MIMO detection problem effectively, this paper proposes to model it as a time series prediction problem, and a bidirectional temporal convolutional network (Bi- TCN) is designed to address it. In Bi- TCN, the encoder extracts the features of the received signal and the channel matrix by applying non-causal dilated convolution, and the decoder outputs the probability distribution of the recovered transmitted signal in parallel. In the experiments, we compare it with traditional and deep learning-based detectors on both i.i.d. and correlated Rayleigh fading channels, respectively. Experimental results empirically demonstrate that Bi- TCN can achieve near-optimal bit-error-rate (BER) performance with considerably low space complexity.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129605593","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 : 2022-10-14DOI: 10.1109/ICIST55546.2022.9926853
Xueli Zhang, Wing W. Y. Ng, Ting Wang
Traffic flow forecasting has been receiving a lot of attention because of its important role in traffic control and management. Accurate traffic forecasting is critical to improving the performance of intelligent transportation systems. However, accurate traffic forecasting still faces the following challenges, including modeling the dynamics of traffic data along the temporal and spatial dimensions, significant differences in peak hour/peak hour traffic, and traffic flow data affected by partial noise. In this paper, we propose a hybrid and robust model with Self-Attention ConvLSTM networks and localized stochastic sensitive (LSS) for traffic flow prediction. The proposed model extracts features with long-range spatiotemporal dependencies with Self-Attention ConvLSTM. To further explore the long-term temporal features, we utilize LSTM module to extract daily and weekly periodic features as assistive features. The LSS reduces sensitivity to unseen samples around training samples and avoids large output fluctuations due to the noise or change of the data. Experiments on real traffic flow datasets show that the proposed method yields better prediction performance compared to other contrast methods.
{"title":"Robust Self-Attention ConvLSTM-based Traffic Flow Prediction Model","authors":"Xueli Zhang, Wing W. Y. Ng, Ting Wang","doi":"10.1109/ICIST55546.2022.9926853","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926853","url":null,"abstract":"Traffic flow forecasting has been receiving a lot of attention because of its important role in traffic control and management. Accurate traffic forecasting is critical to improving the performance of intelligent transportation systems. However, accurate traffic forecasting still faces the following challenges, including modeling the dynamics of traffic data along the temporal and spatial dimensions, significant differences in peak hour/peak hour traffic, and traffic flow data affected by partial noise. In this paper, we propose a hybrid and robust model with Self-Attention ConvLSTM networks and localized stochastic sensitive (LSS) for traffic flow prediction. The proposed model extracts features with long-range spatiotemporal dependencies with Self-Attention ConvLSTM. To further explore the long-term temporal features, we utilize LSTM module to extract daily and weekly periodic features as assistive features. The LSS reduces sensitivity to unseen samples around training samples and avoids large output fluctuations due to the noise or change of the data. Experiments on real traffic flow datasets show that the proposed method yields better prediction performance compared to other contrast methods.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122790059","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 : 2022-10-14DOI: 10.1109/ICIST55546.2022.9926919
Mengxiang Geng, Ming Guo, Jianlong Qiu, Yingchan Cao, Xiangyong Chen
The rapid development of economy and society has led to the rapid increase of the output of logistics outer packaging garbage. How to realize the classification and recycling of logistics packaging garbage by intelligent methods has become a key factor for human beings to achieve sustainable development. To solve this problem, this paper proposes an image recognition model of logistics packaging, which is a convolution neural network model based on multi-scale, and adds channel and spatial attention mechanism. The model uses multi-scale convolution to extract richer image features. The attention mechanism is used to adaptively adjust the parts that need to be focused on, and the feature extraction ability of the model is enhanced. Compared with the traditional manual sorting method, this paper uses the deep learning technology to intelligently and automatically classify the logistics outer packaging. The experimental results show that the classification accuracy of data sets can reach 96% by using the method of deep learning, which is very helpful to improve the classification efficiency of logistics outer packaging.
{"title":"Classification Algorithm of Logistics Packaging Based on Multi-scale Convolutional Neural Network","authors":"Mengxiang Geng, Ming Guo, Jianlong Qiu, Yingchan Cao, Xiangyong Chen","doi":"10.1109/ICIST55546.2022.9926919","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926919","url":null,"abstract":"The rapid development of economy and society has led to the rapid increase of the output of logistics outer packaging garbage. How to realize the classification and recycling of logistics packaging garbage by intelligent methods has become a key factor for human beings to achieve sustainable development. To solve this problem, this paper proposes an image recognition model of logistics packaging, which is a convolution neural network model based on multi-scale, and adds channel and spatial attention mechanism. The model uses multi-scale convolution to extract richer image features. The attention mechanism is used to adaptively adjust the parts that need to be focused on, and the feature extraction ability of the model is enhanced. Compared with the traditional manual sorting method, this paper uses the deep learning technology to intelligently and automatically classify the logistics outer packaging. The experimental results show that the classification accuracy of data sets can reach 96% by using the method of deep learning, which is very helpful to improve the classification efficiency of logistics outer packaging.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125403094","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 this paper, we deal with the weight control of cigarette production process, and we establish a system model with uncertainties and give an event-triggered predictive control algorithm. First, according to the actual situation in the production process, the corresponding relationship between the weight of the cigarette and the height of the leveling disc is established. The uneven distribution and uncertainty of the cut tobacco are taken into account in the model simultaneously. Secondly, a model predictive controller is designed, which can effectively reduce the weight error caused by detection lag, uneven distribution, and random density of cut tobacco. A prediction model and cost function for predicting the future weight of cigarettes are established, and the optimal control sequence is solved when the cost function is the smallest. Finally, considering the limited computing and communication resources of the controller, an event-triggered mechanism is introduced to effectively reduce the computing cost of the controller. Simulation results demonstrate the effectiveness of the proposed event-triggered predictive control method in the cigarette weight control system.
{"title":"An Event-Triggered Predictive Control for Weight Control System","authors":"Xuecheng Zhang, Xiaojie Qiu, Wenchao Meng, Yuliang Li, Lihong Zhang","doi":"10.1109/ICIST55546.2022.9926926","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926926","url":null,"abstract":"In this paper, we deal with the weight control of cigarette production process, and we establish a system model with uncertainties and give an event-triggered predictive control algorithm. First, according to the actual situation in the production process, the corresponding relationship between the weight of the cigarette and the height of the leveling disc is established. The uneven distribution and uncertainty of the cut tobacco are taken into account in the model simultaneously. Secondly, a model predictive controller is designed, which can effectively reduce the weight error caused by detection lag, uneven distribution, and random density of cut tobacco. A prediction model and cost function for predicting the future weight of cigarettes are established, and the optimal control sequence is solved when the cost function is the smallest. Finally, considering the limited computing and communication resources of the controller, an event-triggered mechanism is introduced to effectively reduce the computing cost of the controller. Simulation results demonstrate the effectiveness of the proposed event-triggered predictive control method in the cigarette weight control system.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126239883","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 : 2022-10-14DOI: 10.1109/ICIST55546.2022.9926893
Qilin Ren, Guang-Fu Xue, Xiaoling Gong, Jian Wang
The generation of fuzzy rule base and rule extraction are efficient approaches to enhance the performance of fuzzy rule system. Here, we propose a novel fuzzy neural network structure that can be used for rule extraction. First of all, inspired by the compactly combined fuzzy rule base (CoCo-FRB) and fully combined fuzzy rule base (FuCo-FRB), we develop a new fuzzy rule base, compromise fuzzy rule base (CmPm-FRB), which generates rules by cutting off the long rules and compensating the short ones. In addition, Group Lasso penalty is utilized in the objective function with the rule threshold to produce sparsity in rules in a grouped manner for rule extraction. However, as the Group Lasso penalty is not differentiable at the origin, a tiny bias term is added to the gradient formula to achieve the smoothness. In order to verify the effectiveness of the proposed model, extensive experiments are conducted on ten classification data sets. The empirical results explicitly demonstrate the effectiveness of the proposed model for classification problems.
{"title":"A Novel Fuzzy Rule Based Neuro-system with Sparse Rule Extraction for Classification Problems","authors":"Qilin Ren, Guang-Fu Xue, Xiaoling Gong, Jian Wang","doi":"10.1109/ICIST55546.2022.9926893","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926893","url":null,"abstract":"The generation of fuzzy rule base and rule extraction are efficient approaches to enhance the performance of fuzzy rule system. Here, we propose a novel fuzzy neural network structure that can be used for rule extraction. First of all, inspired by the compactly combined fuzzy rule base (CoCo-FRB) and fully combined fuzzy rule base (FuCo-FRB), we develop a new fuzzy rule base, compromise fuzzy rule base (CmPm-FRB), which generates rules by cutting off the long rules and compensating the short ones. In addition, Group Lasso penalty is utilized in the objective function with the rule threshold to produce sparsity in rules in a grouped manner for rule extraction. However, as the Group Lasso penalty is not differentiable at the origin, a tiny bias term is added to the gradient formula to achieve the smoothness. In order to verify the effectiveness of the proposed model, extensive experiments are conducted on ten classification data sets. The empirical results explicitly demonstrate the effectiveness of the proposed model for classification problems.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114961802","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 : 2022-10-14DOI: 10.1109/ICIST55546.2022.9926830
Gang Baol, Kang Li, Zhenyan Song
This paper introduces a novel kind of discontinu-ous neural networks which are with state-dependent switching external input. The switched external input is defined as a step function with respect to state value. Firstly, we derive a sufficient condition for network state attractivity by dividing the state space according to the swithed external input function and the activation function. At last, one numerical example verifies our results.
{"title":"Attractivity Analysis for Recurrent Neural Networks with State-dependent External Input","authors":"Gang Baol, Kang Li, Zhenyan Song","doi":"10.1109/ICIST55546.2022.9926830","DOIUrl":"https://doi.org/10.1109/ICIST55546.2022.9926830","url":null,"abstract":"This paper introduces a novel kind of discontinu-ous neural networks which are with state-dependent switching external input. The switched external input is defined as a step function with respect to state value. Firstly, we derive a sufficient condition for network state attractivity by dividing the state space according to the swithed external input function and the activation function. At last, one numerical example verifies our results.","PeriodicalId":211213,"journal":{"name":"2022 12th International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123369058","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：