Structured Regression on Multiscale Networks
Authors:
- 27 March 2017
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Intelligent Systems
- Vol. 32 (2) , 23-30
Abstract
Structure-based regression algorithms generally suffer substantive speed losses and have exacting memory requirements compared to their structureless counterparts. Gaussian conditional random field (GCRF) models are one of the most time- and memory-efficient approaches to structured regression. The authors' previous speedups for the GCRF method allow for exact solutions on networks of up to 100,000 nodes and 10 million links. Using multiscale networks, the exact solution for networks of millions of nodes and trillions of links can be solved in a similar amount of time. They walk through the intuitiveness of using multiple scales of networks on a real-life health informatics application. The time and memory demands from using this approach are logarithmic compared to naive implementations.Keywords
This publication has 12 references indexed in Scilit:
- Semi-supervised learning for structured regression on partially observed attributed graphs2015
- Hospital pricing estimation by Gaussian conditional random fields based regression on graphs2014
- Gaussian Conditional Random Fields for Aggregation of Operational Aerosol RetrievalsIEEE Geoscience and Remote Sensing Letters, 2014
- Human symptoms–disease networkNature Communications, 2014
- Neural Gaussian Conditional Random Fields2014
- Continuous Conditional Random Fields for Efficient Regression in Large Fully Connected GraphsProceedings of the AAAI Conference on Artificial Intelligence, 2013
- Modeling Short-Term Energy Load with Continuous Conditional Random FieldsLecture Notes in Computer Science, 2013
- BART: Bayesian additive regression treesThe Annals of Applied Statistics, 2010
- Boosting Algorithms: Regularization, Prediction and Model FittingStatistical Science, 2007
- Laplacian graph eigenvectorsLinear Algebra and its Applications, 1998