2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)最新文献
Pub Date : 2019-10-01DOI: 10.1109/IUCC/DSCI/SmartCNS.2019.00163
Di Yang, Huamin Yang, Peng Wang, Songjiang Li
Multi-step traffic flow prediction extends short-term single-step prediction to long-term prediction, which is more significant in many basic application in intelligent transportation systems, such as traffic planning. A main problem of multi-step prediction is that the error accumulation as steps increase, resulting in prediction performance degradation. In this work, combining recursive and multi-output strategies, we proposed a deep learning model, named MARNN, for multi-step traffic flow prediction. Specifically, we jointly consider recurrent neural network as dynamic neural network for simulating the dynamic characteristics in traffic time series as recursive strategy does and multi-output strategy for decreasing the accumulated error as step increases. In addition, we introduce attention mechanism for adaptively seeking correlated important information among traffic time series to improve prediction performance. The experiments on real traffic data show the advantages of MARNN model over other four baseline models, demonstrating the potential and promising capability of the proposed model on multi-step traffic flow prediction.
{"title":"Multi-Step Traffic Flow Prediction Using Recurrent Neural Network","authors":"Di Yang, Huamin Yang, Peng Wang, Songjiang Li","doi":"10.1109/IUCC/DSCI/SmartCNS.2019.00163","DOIUrl":"https://doi.org/10.1109/IUCC/DSCI/SmartCNS.2019.00163","url":null,"abstract":"Multi-step traffic flow prediction extends short-term single-step prediction to long-term prediction, which is more significant in many basic application in intelligent transportation systems, such as traffic planning. A main problem of multi-step prediction is that the error accumulation as steps increase, resulting in prediction performance degradation. In this work, combining recursive and multi-output strategies, we proposed a deep learning model, named MARNN, for multi-step traffic flow prediction. Specifically, we jointly consider recurrent neural network as dynamic neural network for simulating the dynamic characteristics in traffic time series as recursive strategy does and multi-output strategy for decreasing the accumulated error as step increases. In addition, we introduce attention mechanism for adaptively seeking correlated important information among traffic time series to improve prediction performance. The experiments on real traffic data show the advantages of MARNN model over other four baseline models, demonstrating the potential and promising capability of the proposed model on multi-step traffic flow prediction.","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131828360","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 : 2019-10-01DOI: 10.1109/IUCC/DSCI/SmartCNS.2019.00148
Tengyuan Liu, Ruifang Li
With the development of vehicle network services and related software, the dedicated short-range communication (DSRC) frequency band is not sufficient to carry the increased wireless service demand in the in-vehicle network. The large television spectrum (i.e., the TV white space spectrum) issued by the US Federal Communications Commission for cognitive access will be used to provide additional bandwidth to the in-vehicle network, and the effective channel allocation scheme for TV White Space (TVWS) has become our goal. In this paper, we design a channel allocation scheme based on the Hungarian algorithm. Our main goal is to increase network throughput while minimizing network latency. We present simulation schemes and numerical evaluations to illustrate the desired performance of the proposed channel allocation scheme.
{"title":"VANET Channel Allocation Scheme Based on Hungarian Algorithm","authors":"Tengyuan Liu, Ruifang Li","doi":"10.1109/IUCC/DSCI/SmartCNS.2019.00148","DOIUrl":"https://doi.org/10.1109/IUCC/DSCI/SmartCNS.2019.00148","url":null,"abstract":"With the development of vehicle network services and related software, the dedicated short-range communication (DSRC) frequency band is not sufficient to carry the increased wireless service demand in the in-vehicle network. The large television spectrum (i.e., the TV white space spectrum) issued by the US Federal Communications Commission for cognitive access will be used to provide additional bandwidth to the in-vehicle network, and the effective channel allocation scheme for TV White Space (TVWS) has become our goal. In this paper, we design a channel allocation scheme based on the Hungarian algorithm. Our main goal is to increase network throughput while minimizing network latency. We present simulation schemes and numerical evaluations to illustrate the desired performance of the proposed channel allocation scheme.","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"71 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130784402","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 : 2019-10-01DOI: 10.1109/iucc/dsci/smartcns.2019.00015
{"title":"Message from the EMCA 2019 Workshop Chairs","authors":"","doi":"10.1109/iucc/dsci/smartcns.2019.00015","DOIUrl":"https://doi.org/10.1109/iucc/dsci/smartcns.2019.00015","url":null,"abstract":"","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133031925","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}
Recently, Machine Learning methods have shown great improvement while analyzing the biomedical data. Raman Spectroscopy (RS), a non-invasive technique, and widely used in screening to diagnose the oral cancer. In order to spot cancer in a smarter and faster way, we have employed Meta-Learning (ML) techniques to learn such as Bagging and Boosting on RS data. Further, we employed normal and tumor tissue class classification by Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA) and Adaptive Boosting (AdaBoost) classifiers. The present study aims at examining the RS data with total 110 samples, including 57 tumor and 53 normal ones. To evaluate the performance, we have used the training samples to optimize, and testing samples to generalize the model parameters. The results show that the AdaBoost classifier with Bagging techniques showed the significant changes in accuracy.
{"title":"Meta-Learning Techniques to Analyze the Raman Data for Optical Diagnosis of Oral Cancer Detection","authors":"Mukta Sharma, Lokesh Sharma, M. Jeng, Liann-Be Chang, Shiang-Fu Huang, Shih-Lin Wu","doi":"10.1109/IUCC/DSCI/SmartCNS.2019.00134","DOIUrl":"https://doi.org/10.1109/IUCC/DSCI/SmartCNS.2019.00134","url":null,"abstract":"Recently, Machine Learning methods have shown great improvement while analyzing the biomedical data. Raman Spectroscopy (RS), a non-invasive technique, and widely used in screening to diagnose the oral cancer. In order to spot cancer in a smarter and faster way, we have employed Meta-Learning (ML) techniques to learn such as Bagging and Boosting on RS data. Further, we employed normal and tumor tissue class classification by Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA) and Adaptive Boosting (AdaBoost) classifiers. The present study aims at examining the RS data with total 110 samples, including 57 tumor and 53 normal ones. To evaluate the performance, we have used the training samples to optimize, and testing samples to generalize the model parameters. The results show that the AdaBoost classifier with Bagging techniques showed the significant changes in accuracy.","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115738565","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 : 2019-10-01DOI: 10.1109/IUCC/DSCI/SmartCNS.2019.00109
Haixiao Li, Dong Yu
Node Localization in Wireless Sensor Network (WSN) is the process of localizing the unknown node in use of anchor nodes' coordinates which are known in advance or can be got by GPS. As one of typical range-free localization algorithms, Distance Vector Hop (DV-Hop) algorithm is low-cost and doesn't require for additional hardware for ranging. However, the localization accuracy of DV-Hop algorithm is low because as parameters in localization algorithm the distances between the unknown node and anchor nodes are estimated by multiplying number of hops and average hop distance. The error in estimated distance leads to error accumulation in localization result. Therefore DV-Hop algorithm is usually applied under circumstances with low requirement for localization accuracy. In view of the main reason of the error in DV-Hop algorithm, an improved algorithm i.e. DV-Hop algorithm based on Error Vector Modification (EVM-DV-Hop) is proposed in this paper. The improved algorithm reduces the localization error by modifying the estimated coordinate of the unknown node with the help of Error Vector determined by anchor nodes with a certain degree of similarity to the unknown node. It is shown in simulation results that the improved algorithm reduces the localization error and raise the localization accuracy of DV-Hop algorithm effectively.
{"title":"An Improved DV-Hop Localization Algorithm Based on Error Vector Modification for Wireless Sensor Network","authors":"Haixiao Li, Dong Yu","doi":"10.1109/IUCC/DSCI/SmartCNS.2019.00109","DOIUrl":"https://doi.org/10.1109/IUCC/DSCI/SmartCNS.2019.00109","url":null,"abstract":"Node Localization in Wireless Sensor Network (WSN) is the process of localizing the unknown node in use of anchor nodes' coordinates which are known in advance or can be got by GPS. As one of typical range-free localization algorithms, Distance Vector Hop (DV-Hop) algorithm is low-cost and doesn't require for additional hardware for ranging. However, the localization accuracy of DV-Hop algorithm is low because as parameters in localization algorithm the distances between the unknown node and anchor nodes are estimated by multiplying number of hops and average hop distance. The error in estimated distance leads to error accumulation in localization result. Therefore DV-Hop algorithm is usually applied under circumstances with low requirement for localization accuracy. In view of the main reason of the error in DV-Hop algorithm, an improved algorithm i.e. DV-Hop algorithm based on Error Vector Modification (EVM-DV-Hop) is proposed in this paper. The improved algorithm reduces the localization error by modifying the estimated coordinate of the unknown node with the help of Error Vector determined by anchor nodes with a certain degree of similarity to the unknown node. It is shown in simulation results that the improved algorithm reduces the localization error and raise the localization accuracy of DV-Hop algorithm effectively.","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"114 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117214571","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 : 2019-10-01DOI: 10.1109/IUCC/DSCI/SmartCNS.2019.00063
H. Yin, Yurong Wei
As an essential method of fire prevention and disaster control, smoke detection is of great significance to production and life. At present, the convolutional neural network (CNN) has achieved good results in the research of smoke detection. However, the detection accuracy is not high for some scenes. For example, the wind speed is tremendous, and the shape of the smoke changes rapidly. In order to deal with this problem better, this paper proposes an improved algorithm based on cascading classification and deep convolutional neural network. In the cascading classification part, we improve the cascading structure and make it select the appropriate parameter threshold for the smoke generated in different scenes. The convolutional neural network structure is trained to extract the variation characteristics of smoke better. Also, we optimize the parameters on the target data set. The experimental results show that the algorithm has achieved excellent results in accuracy and speed on the selected smoke detection data sets.
{"title":"An Improved Algorithm Based on Convolutional Neural Network for Smoke Detection","authors":"H. Yin, Yurong Wei","doi":"10.1109/IUCC/DSCI/SmartCNS.2019.00063","DOIUrl":"https://doi.org/10.1109/IUCC/DSCI/SmartCNS.2019.00063","url":null,"abstract":"As an essential method of fire prevention and disaster control, smoke detection is of great significance to production and life. At present, the convolutional neural network (CNN) has achieved good results in the research of smoke detection. However, the detection accuracy is not high for some scenes. For example, the wind speed is tremendous, and the shape of the smoke changes rapidly. In order to deal with this problem better, this paper proposes an improved algorithm based on cascading classification and deep convolutional neural network. In the cascading classification part, we improve the cascading structure and make it select the appropriate parameter threshold for the smoke generated in different scenes. The convolutional neural network structure is trained to extract the variation characteristics of smoke better. Also, we optimize the parameters on the target data set. The experimental results show that the algorithm has achieved excellent results in accuracy and speed on the selected smoke detection data sets.","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"445 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125768585","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 : 2019-10-01DOI: 10.1109/iucc/dsci/smartcns.2019.00002
{"title":"Title Page III","authors":"","doi":"10.1109/iucc/dsci/smartcns.2019.00002","DOIUrl":"https://doi.org/10.1109/iucc/dsci/smartcns.2019.00002","url":null,"abstract":"","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125829933","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 : 2019-10-01DOI: 10.1109/IUCC/DSCI/SmartCNS.2019.00156
Yao Zhou, Yiying Niu, Yi Feng
5G is the research and development focus of global communication technology, and is an important engine for promoting the development of the national economy and improving the level of information. As the basic strategic resource of wireless communication, spectrum is crucial to the development of 5G industry. In order to better develop the IMT-2020 system, the WRC-15 Conference resolved to identify WRC-19 1.13 issues: deliberating to identify frequency bands for future development of International Mobile Telecommunications (IMT), including the possibility of making additional divisions for mobile services as the main business. This paper is based on the WRC-19 1.13 issue, the 24.75-27.5 GHz band as its key frequency band, which is the main research content of this paper. It mainly deals with the study of adjacent frequency interference of IMT-2020 system to the Earth exploration satellite service (EESS) (passive) in 24.75-27.5 GHz frequency band.
{"title":"Sharing and Compatibility Studies of Eess(Passive) and IMT System in 24.25-27.5 GHz","authors":"Yao Zhou, Yiying Niu, Yi Feng","doi":"10.1109/IUCC/DSCI/SmartCNS.2019.00156","DOIUrl":"https://doi.org/10.1109/IUCC/DSCI/SmartCNS.2019.00156","url":null,"abstract":"5G is the research and development focus of global communication technology, and is an important engine for promoting the development of the national economy and improving the level of information. As the basic strategic resource of wireless communication, spectrum is crucial to the development of 5G industry. In order to better develop the IMT-2020 system, the WRC-15 Conference resolved to identify WRC-19 1.13 issues: deliberating to identify frequency bands for future development of International Mobile Telecommunications (IMT), including the possibility of making additional divisions for mobile services as the main business. This paper is based on the WRC-19 1.13 issue, the 24.75-27.5 GHz band as its key frequency band, which is the main research content of this paper. It mainly deals with the study of adjacent frequency interference of IMT-2020 system to the Earth exploration satellite service (EESS) (passive) in 24.75-27.5 GHz frequency band.","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125902544","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}
Continuous query processing over sliding window is an important problem in stream data management. Given the sliding window W and the continuous query q, q monitors to the multidimensional data objects in the window. When the window slides, q returns all the skyline objects in the window. Many scholars have carried out researches on such problems. The core idea is to delete objects that cannot be query results by using the temporal sequence relationship between objects in the window, and when the window slides, the algorithm can find the query results from the rest. However, the algorithm is sensitive to data timing relationships such as problems. In the worst case, the size of the candidate object equals to the size of the data in the window. In this paper, we propose a partition-based framework to support continuous skyline query over sliding window. It partitions the window into a group of sub-window, and maintain the skyline objects in each sub-window. In this way, it could effectively overcome the impact the object arrived order to the algorithm performance. In addition, we propose a self-adaptive algorithm to partition the window according to the distribution of streaming data. A large number of experiments prove the effectiveness and high efficiency of the proposed algorithm.
{"title":"Continuous Skyline Query Processing Algorithm Based on Sharding Technology Under Sliding Window Model","authors":"Xiufeng Xia, T. Yu, Rui Zhu, Jiajia Li, Xiangyu Liu, Chuanyu Zong","doi":"10.1109/iucc/dsci/smartcns.2019.00036","DOIUrl":"https://doi.org/10.1109/iucc/dsci/smartcns.2019.00036","url":null,"abstract":"Continuous query processing over sliding window is an important problem in stream data management. Given the sliding window W and the continuous query q, q monitors to the multidimensional data objects in the window. When the window slides, q returns all the skyline objects in the window. Many scholars have carried out researches on such problems. The core idea is to delete objects that cannot be query results by using the temporal sequence relationship between objects in the window, and when the window slides, the algorithm can find the query results from the rest. However, the algorithm is sensitive to data timing relationships such as problems. In the worst case, the size of the candidate object equals to the size of the data in the window. In this paper, we propose a partition-based framework to support continuous skyline query over sliding window. It partitions the window into a group of sub-window, and maintain the skyline objects in each sub-window. In this way, it could effectively overcome the impact the object arrived order to the algorithm performance. In addition, we propose a self-adaptive algorithm to partition the window according to the distribution of streaming data. A large number of experiments prove the effectiveness and high efficiency of the proposed algorithm.","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129928016","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 : 2019-10-01DOI: 10.1109/iucc/dsci/smartcns.2019.00011
{"title":"Message from the IMHCS 2019 Workshop Chairs","authors":"","doi":"10.1109/iucc/dsci/smartcns.2019.00011","DOIUrl":"https://doi.org/10.1109/iucc/dsci/smartcns.2019.00011","url":null,"abstract":"","PeriodicalId":410905,"journal":{"name":"2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130182177","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}
2019 IEEE International Conferences on Ubiquitous Computing & Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS)
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：