{"title":"基于改进马尔可夫模型的无线用户缺失位置移动预测","authors":"Junyao Guo, Lu Liu, Sihai Zhang, Jinkang Zhu","doi":"10.1109/BigDataCongress.2019.00031","DOIUrl":null,"url":null,"abstract":"Mobility prediction is an interesting topic attracting many researchers and both prediction theory and models are explored in the existing literature. The entropy metric to evaluate the mobility predictability of individuals gives a theoretical upper bound and lower bound of prediction probability, although the achieved accuracies of users with the same predictability vary. In this work, we investigate the missing locations phenomenon which means the users visit new locations in the testing set. The major difference of theoretical bound between with and without missing locations are found, which shows that users without missing locations are easier to predict. After discussing the impact of missing locations on the prediction accuracy, a modified Markov chain prediction model is proposed to deal with the presence of missing positions. Finally, the correlation between accuracy and predictability can be modeled as the Gaussian distribution and the standard deviation modeled with missing locations can be modeled as double Gaussian function, while that without missing locations can be modeled as the third-order polynomial function.","PeriodicalId":335850,"journal":{"name":"2019 IEEE International Congress on Big Data (BigDataCongress)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Mobility Prediction with Missing Locations Based on Modified Markov Model for Wireless Users\",\"authors\":\"Junyao Guo, Lu Liu, Sihai Zhang, Jinkang Zhu\",\"doi\":\"10.1109/BigDataCongress.2019.00031\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Mobility prediction is an interesting topic attracting many researchers and both prediction theory and models are explored in the existing literature. The entropy metric to evaluate the mobility predictability of individuals gives a theoretical upper bound and lower bound of prediction probability, although the achieved accuracies of users with the same predictability vary. In this work, we investigate the missing locations phenomenon which means the users visit new locations in the testing set. The major difference of theoretical bound between with and without missing locations are found, which shows that users without missing locations are easier to predict. After discussing the impact of missing locations on the prediction accuracy, a modified Markov chain prediction model is proposed to deal with the presence of missing positions. Finally, the correlation between accuracy and predictability can be modeled as the Gaussian distribution and the standard deviation modeled with missing locations can be modeled as double Gaussian function, while that without missing locations can be modeled as the third-order polynomial function.\",\"PeriodicalId\":335850,\"journal\":{\"name\":\"2019 IEEE International Congress on Big Data (BigDataCongress)\",\"volume\":\"22 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2019-07-08\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2019 IEEE International Congress on Big Data (BigDataCongress)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/BigDataCongress.2019.00031\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE International Congress on Big Data (BigDataCongress)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/BigDataCongress.2019.00031","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Mobility Prediction with Missing Locations Based on Modified Markov Model for Wireless Users
Mobility prediction is an interesting topic attracting many researchers and both prediction theory and models are explored in the existing literature. The entropy metric to evaluate the mobility predictability of individuals gives a theoretical upper bound and lower bound of prediction probability, although the achieved accuracies of users with the same predictability vary. In this work, we investigate the missing locations phenomenon which means the users visit new locations in the testing set. The major difference of theoretical bound between with and without missing locations are found, which shows that users without missing locations are easier to predict. After discussing the impact of missing locations on the prediction accuracy, a modified Markov chain prediction model is proposed to deal with the presence of missing positions. Finally, the correlation between accuracy and predictability can be modeled as the Gaussian distribution and the standard deviation modeled with missing locations can be modeled as double Gaussian function, while that without missing locations can be modeled as the third-order polynomial function.