Statistical analysis of base data from the wsr-88d weather radar and description of mesocyclone features using radial basis function networks

摘要

The classification power of these neural nets can be improved with more extensive training with a larger set of training patches containing circulations. They show good potential in controlling false alarms. These can be tested with more real time data and benchmarked against existing algorithms and could be made as part of the Doppler weather radar algorithms in the future. We are currently investigating support vector machines as an alternative to radial basis function networks.