Evolving Spatially Aggregated Features from Satellite Imagery for Regional Modeling

Satellite imagery and remote sensing provide explanatory variables at relatively high resolutions for modeling geospatial phenomena, yet regional summaries are often desirable for analysis and actionable insight. In this paper, we propose a novel method of inducing spatial aggregations as a component of the machine learning process, yielding regional model features whose construction is driven by model prediction performance rather than prior assumptions. Our results demonstrate that Genetic Programming is particularly well suited to this type of feature construction because it can automatically synthesize appropriate aggregations, as well as better incorporate them into predictive models compared to other regression methods we tested. In our experiments we consider a specific problem instance and real-world dataset relevant to predicting snow properties in high-mountain Asia.

[1]  Thomas H. Painter,et al.  Retrieval of subpixel snow covered area, grain size, and albedo from MODIS , 2009 .

[2]  Albert Rango,et al.  Areal distribution of snow water equivalent evaluated by snow cover monitoring , 1981 .

[3]  D. Wolfe,et al.  Nonparametric Statistical Methods. , 1974 .

[4]  Josh Bongard,et al.  Inductive machine learning for improved estimation of catchment-scale snow water equivalent , 2015 .

[5]  Krzysztof Krawiec,et al.  Genetic Programming-based Construction of Features for Machine Learning and Knowledge Discovery Tasks , 2002, Genetic Programming and Evolvable Machines.

[6]  A. E. Hoerl,et al.  Ridge regression: biased estimation for nonorthogonal problems , 2000 .

[7]  R. Tibshirani.  Regression Shrinkage and Selection via the Lasso , 1996 .

[8]  Parag S. Narvekar,et al.  Assessment of the NASA AMSR-E SWE Product , 2010, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[9]  W. Tobler.  A Computer Movie Simulating Urban Growth in the Detroit Region , 1970 .

[10]  A. E. Hoerl,et al.  Ridge Regression: Applications to Nonorthogonal Problems , 1970 .

[11]  John Rees,et al.  Regional modelling of geohazard change , 2009 .

[12]  Hod Lipson,et al.  Automated reverse engineering of nonlinear dynamical systems , 2007, Proceedings of the National Academy of Sciences.

[13]  John R. Koza,et al.  Genetic programming - on the programming of computers by means of natural selection , 1993, Complex adaptive systems.

[14]  Michael D. Schmidt,et al.  Automated refinement and inference of analytical models for metabolic networks , 2011, Physical biology.

[15]  Paul R. Houser,et al.  Factors affecting remotely sensed snow water equivalent uncertainty , 2005 .

[16]  Thomas H. Painter,et al.  Time-space continuity of daily maps of fractional snow cover and albedo from MODIS , 2008 .

[17]  Hod Lipson,et al.  Age-fitness pareto optimization , 2010, GECCO '10.

[18]  Krzysztof Krawiec,et al.  Genetic Programming for Estimation of Heat Flux between the Atmosphere and Sea Ice in Polar Regions , 2015, GECCO.