A. Grover, A. Kapoor, and E. Horvitz. Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, page 379--386. New York, NY, USA, ACM, (2015)
DOI: 10.1145/2783258.2783275
Abstract
Weather forecasting is a canonical predictive challenge that has depended primarily on model-based methods. We explore new directions with forecasting weather as a data-intensive challenge that involves inferences across space and time. We study specifically the power of making predictions via a hybrid approach that combines discriminatively trained predictive models with a deep neural network that models the joint statistics of a set of weather-related variables. We show how the base model can be enhanced with spatial interpolation that uses learned long-range spatial dependencies. We also derive an efficient learning and inference procedure that allows for large scale optimization of the model parameters. We evaluate the methods with experiments on real-world meteorological data that highlight the promise of the approach.
%0 Conference Paper
%1 Grover:2015:DHM:2783258.2783275
%A Grover, Aditya
%A Kapoor, Ashish
%A Horvitz, Eric
%B Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
%C New York, NY, USA
%D 2015
%I ACM
%K deep_learning environment neural_network
%P 379--386
%R 10.1145/2783258.2783275
%T A Deep Hybrid Model for Weather Forecasting
%U http://doi.acm.org/10.1145/2783258.2783275
%X Weather forecasting is a canonical predictive challenge that has depended primarily on model-based methods. We explore new directions with forecasting weather as a data-intensive challenge that involves inferences across space and time. We study specifically the power of making predictions via a hybrid approach that combines discriminatively trained predictive models with a deep neural network that models the joint statistics of a set of weather-related variables. We show how the base model can be enhanced with spatial interpolation that uses learned long-range spatial dependencies. We also derive an efficient learning and inference procedure that allows for large scale optimization of the model parameters. We evaluate the methods with experiments on real-world meteorological data that highlight the promise of the approach.
%@ 978-1-4503-3664-2
@inproceedings{Grover:2015:DHM:2783258.2783275,
abstract = {Weather forecasting is a canonical predictive challenge that has depended primarily on model-based methods. We explore new directions with forecasting weather as a data-intensive challenge that involves inferences across space and time. We study specifically the power of making predictions via a hybrid approach that combines discriminatively trained predictive models with a deep neural network that models the joint statistics of a set of weather-related variables. We show how the base model can be enhanced with spatial interpolation that uses learned long-range spatial dependencies. We also derive an efficient learning and inference procedure that allows for large scale optimization of the model parameters. We evaluate the methods with experiments on real-world meteorological data that highlight the promise of the approach.},
acmid = {2783275},
added-at = {2016-09-29T15:22:19.000+0200},
address = {New York, NY, USA},
author = {Grover, Aditya and Kapoor, Ashish and Horvitz, Eric},
biburl = {https://www.bibsonomy.org/bibtex/2244b3d0ea886d75ec5db548d3aa2db59/dallmann},
booktitle = {Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining},
doi = {10.1145/2783258.2783275},
interhash = {46bc7aa2bf55beeb0b597815054a6434},
intrahash = {244b3d0ea886d75ec5db548d3aa2db59},
isbn = {978-1-4503-3664-2},
keywords = {deep_learning environment neural_network},
location = {Sydney, NSW, Australia},
numpages = {8},
pages = {379--386},
publisher = {ACM},
series = {KDD '15},
timestamp = {2016-09-29T15:22:19.000+0200},
title = {A Deep Hybrid Model for Weather Forecasting},
url = {http://doi.acm.org/10.1145/2783258.2783275},
year = 2015
}