Online Program Home
My Program

Abstract Details

Activity Number: 341
Type: Topic Contributed
Date/Time: Tuesday, August 2, 2016 : 10:30 AM to 12:20 PM
Sponsor: Biometrics Section
Abstract #318814 View Presentation
Title: Functional Nearest Neighbor Ensembles with Application to Water Quality Monitoring
Author(s): Jan Gertheiss* and Karen Fuchs
Companies: Clausthal University of Technology and Siemens AG
Keywords: functional data ; nearest neighbors ; nonparametric classification ; ensemble method ; variable selection ; feature selection

We consider data from an experiment where measurements obtained longitudinally from various cell-based sensor chips are used to detect paracetamol added to an nutrient medium. For analyzing the data, we propose an ensemble method for nonparametric classification with functional data that inherently provides automatic and interpretable feature selection. It is designed for single as well as multiple functional (and non-functional) covariates. The ensemble members are posterior probability estimates which are obtained using k-nearest-neighbors based on different semi-metrics, with each of those semi-metrics focusing on a specific curve feature. Each ensemble member, and thus each curve feature, is weighted by a specific coefficient which is estimated using a proper scoring rule with implicit lasso-type penalty, such that some coefficients can be estimated to be exactly zero. Thus, the ensemble automatically provides feature (and variable) selection.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association