TY - GEN
T1 - Tornado detection with Kernel-based classifiers from WSR-88D radar data
AU - Trafalis, Theodore B.
AU - Santosa, Budi
AU - Richman, Michael B.
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2016.
PY - 2016
Y1 - 2016
N2 - Detection of tornadoes that provides warning times sufficient for evasive action prior to a tornado strike has been a well-established objective of weather forecasters. With modern technology, progress has been made on increasing the average lead time of such warnings, which translates into a number of lives saved. Recently, machine learning (e.g., kernel methods) has been added to the collection of techniques brought to bear on severe weather prediction. In this chapter, we seek to extend this innovation by introducing and applying two types of kernel-based methods, support vector machines and minimax probability machines to detect tornadoes, using attributes from radar derived velocity data. These two approaches utilize kernel methods to address nonlinearity of the data in the input space. The approaches are based on maximizing the margin between two different classes: tornado and no tornado. The use of the Weather Surveillance Radar 1988 Doppler, with continuous data streaming every 6min, presents a source for a dynamic data driven application system. The results are compared to those produced by neural networks (NN). Findings indicate that these kernel approaches are significantly more accurate than NN for the tornado detection problem.
AB - Detection of tornadoes that provides warning times sufficient for evasive action prior to a tornado strike has been a well-established objective of weather forecasters. With modern technology, progress has been made on increasing the average lead time of such warnings, which translates into a number of lives saved. Recently, machine learning (e.g., kernel methods) has been added to the collection of techniques brought to bear on severe weather prediction. In this chapter, we seek to extend this innovation by introducing and applying two types of kernel-based methods, support vector machines and minimax probability machines to detect tornadoes, using attributes from radar derived velocity data. These two approaches utilize kernel methods to address nonlinearity of the data in the input space. The approaches are based on maximizing the margin between two different classes: tornado and no tornado. The use of the Weather Surveillance Radar 1988 Doppler, with continuous data streaming every 6min, presents a source for a dynamic data driven application system. The results are compared to those produced by neural networks (NN). Findings indicate that these kernel approaches are significantly more accurate than NN for the tornado detection problem.
KW - Dynamic data driven application
KW - Feedforward neural networks
KW - Generalization error
KW - Kernel methods
KW - Tornado detection
UR - http://www.scopus.com/inward/record.url?scp=85006043192&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-43709-5_16
DO - 10.1007/978-3-319-43709-5_16
M3 - Conference contribution
AN - SCOPUS:85006043192
SN - 9783319437071
T3 - Springer Proceedings in Mathematics and Statistics
SP - 329
EP - 344
BT - Dynamics of Disasters–Key Concepts, Models, Algorithms, and Insights, 2015
A2 - Nagurney, Anna
A2 - Kotsireas, Ilias S.
A2 - Pardalos, Panos M.
PB - Springer New York LLC
T2 - 2nd International Conference on Dynamics of Disasters, 2015
Y2 - 29 June 2015 through 2 July 2015
ER -