Covariate Shift Adaptation via Sparse Filtering for High-Dimensional Periodic Data

Fabio Massimo Zennaro, Ke ChenResearch 2016

Abstract

Let us consider a semi-supervised setting in which we are given the training data X tr , the corresponding labels Y tr defining C classes, and the test data X tst , with the assumptions of covariate shift and periodic data structure.To devise a SF-based algorithm that performs CSA in this setting, we need to enrich the original SF algorithm with two additional properties: (i) the ability to generate
BibTeX
@inproceedings{zennaro2016covariate,
  author = {Fabio Massimo Zennaro and Ke Chen},
  title = {{Covariate Shift Adaptation via Sparse Filtering for High-Dimensional Periodic Data}},
  booktitle = {Research},
  year = {2016},
}

Loading related content...

Figure