Improving partial mutual information-based input variable selection by consideration of boundary issues associated with bandwidth estimation

Li, X.Zecchin, A.Maier, H.2015-11-272015-11-272015Environmental Modelling and Software, 2015; 71:78-961364-81521873-6726http://hdl.handle.net/2440/96957Abstract not availableen© 2015 Elsevier Ltd. All rights reserved.Artificial neural networks; Data-driven models; Partial mutual information; Kernel density estimation; Kernel bandwidth; Boundary issues; Hydrology and water resources; Input variable selectionImproving partial mutual information-based input variable selection by consideration of boundary issues associated with bandwidth estimationJournal article003003197610.1016/j.envsoft.2015.05.0130003586275000072-s2.0-84934975077194271Zecchin, A. [0000-0001-8908-7023]Maier, H. [0000-0002-0277-6887]