Enhanced spatiotemporal relational probability trees and forests

Abstract

Many real world domains are inherently spatiotemporal in nature. In this work, we introduce significant enhancements to two spatiotemporal relational learning methods, the spatiotemporal relational probability tree and the spatiotemporal relational random forest, that increase their ability to learn using spatiotemporal data. We enabled the models to formulate questions on both objects and the scalar and vector fields within and around objects, allowing the models to differentiate based on the gradient, divergence, and curl and to recognize the shape of point clouds defined by fields. This enables the model to ask questions about the change of a shape over time or about its orientation. These additions are validated on several real-world hazardous weather datasets. We demonstrate that these additions enable the models to learn robust classifiers that outperform the versions without these new additions. In

Spatial Structured Prediction Models: Applications, Challenges, and Techniques

“…Spatial relationship features can be readily used in rule-based or decision tree-based models. Examples of techniques include spatiotemporal probability tree model to classify meteorological data on storms [36], [37], multi-relational spatial classification [38], prediction based on spatial association rules [39].…”

Enhancing understanding and improving prediction of severe weather through spatiotemporal relational learning

“…We have previously introduced Spatiotemporal Relational Probability Trees (SRPT) and their associated ensemble forests (SRRF) (McGovern et al 2008 , 2010 , 2011 , 2013 ). In this paper, we focus on the new aspects.…”

“…We call these fielded objects , following the convention in geographic information systems (Goodchild et al 2007 ; Cova and Goodchild 2002 ). We have previously described our modifications in McGovern et al ( 2010 , 2011 , 2013 ) and we briefly describe the data through an example here.…”

“…In both cases, they take on a variety of shapes but they rarely take on a canonical shape such as a circle, cylinder, or cone. In previous work, we had implemented a shape recognition algorithm for such shapes (McGovern et al 2013 ) but it limited the identifiable shapes.…”

“…SRL has proven to be very successful in a wide variety of applications (e.g., Neville et al 2005 ; Fast et al 2007 ; Neville and Jensen 2007 ; Raghavan et al 2012 ). We previously developed the Spatiotemporal Relational Probability Tree (SRPT) and its related Spatiotemporal Relational Random Forest (SRRF) techniques (McGovern et al 2008 , 2010 , 2011 , 2013 ) and have demonstrated that they can be successfully applied to severe weather applications.…”