English

Multiresolution Tensor Learning for Efficient and Interpretable Spatial Analysis

Machine Learning 2020-08-18 v5 Machine Learning

Abstract

Efficient and interpretable spatial analysis is crucial in many fields such as geology, sports, and climate science. Tensor latent factor models can describe higher-order correlations for spatial data. However, they are computationally expensive to train and are sensitive to initialization, leading to spatially incoherent, uninterpretable results. We develop a novel Multiresolution Tensor Learning (MRTL) algorithm for efficiently learning interpretable spatial patterns. MRTL initializes the latent factors from an approximate full-rank tensor model for improved interpretability and progressively learns from a coarse resolution to the fine resolution to reduce computation. We also prove the theoretical convergence and computational complexity of MRTL. When applied to two real-world datasets, MRTL demonstrates 4~5x speedup compared to a fixed resolution approach while yielding accurate and interpretable latent factors.

Keywords

Cite

@article{arxiv.2002.05578,
  title  = {Multiresolution Tensor Learning for Efficient and Interpretable Spatial Analysis},
  author = {Jung Yeon Park and Kenneth Theo Carr and Stephan Zheng and Yisong Yue and Rose Yu},
  journal= {arXiv preprint arXiv:2002.05578},
  year   = {2020}
}

Comments

ICML 2020