Collective Inference for Handling Autocorrelation in Network Regression
作者： Corrado Loglisci , Annalisa Appice , Donato Malerba
DOI: 10.1007/978-3-319-08326-1_58
关键词: Regression 、 Predictive inference 、 Computer science 、 Iterative method 、 Data mining 、 Task (computing) 、 Variables 、 Inference 、 Artificial intelligence 、 Node (networking) 、 Machine learning 、 Autocorrelation
摘要: In predictive data mining tasks, we should account for autocorrelations of both the independent variables and dependent variable, which can observe in neighborhood a target node that same node. The prediction on be based value neighbours might even unavailable. To address this problem, values inferred collectively. We present novel computational solution to perform collective inferences network regression task. define an iterative algorithm, order make about predictions multiple nodes simultaneously feed back more reliable made by previous models labeled network. Experiments investigate effectiveness proposed algorithm spatial networks.
springer.com
sci-hub.st 参考文章(5)
Luís Torgo, Orlando Ohashi, Wind speed forecasting using spatio-temporal indicators european conference on artificial intelligence. pp. 975- 980 ,(2012) , 10.3233/978-1-61499-098-7-975
Annalisa Appice, Saso Džeroski, Stepwise Induction of Multi-target Model Trees european conference on machine learning. pp. 502- 509 ,(2007) , 10.1007/978-3-540-74958-5_46
Daniela Stojanova, Michelangelo Ceci, Annalisa Appice, Sašo Džeroski, Network regression with predictive clustering trees Data Mining and Knowledge Discovery. ,vol. 25, pp. 378- 413 ,(2012) , 10.1007/S10618-012-0278-6
Arthur Getis, A History of the Concept of Spatial Autocorrelation: A Geographer's Perspective Geographical Analysis. ,vol. 40, pp. 297- 309 ,(2008) , 10.1111/J.1538-4632.2008.00727.X
Machine Learning: ECML 2007 Springer Berlin Heidelberg. ,(2007) , 10.1007/978-3-540-74958-5