|
Activity Number:
|
72
- Methods for Extreme Values in Environmental Data
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
|
Abstract #313002
|
|
|
Title:
|
Tornado: Classification, Correlation, Prediction
|
|
Author(s):
|
Thilini Vasana Mahanama* and Dimitri Volchenkov
|
|
Companies:
|
Texas Tech University and Texas Tech University
|
|
Keywords:
|
Non-negative Matrix Factorization;
Copula;
Long Short-Term Memory Neural Networks
|
|
Abstract:
|
National Oceanic and Atmospheric Administration (NOAA) annually reports around 1,300 tornado events hitting the US soil. Non-negative matrix factorization is used to classify tornado events with the account for property losses. The results of linear regression about that property losses are roughly proportional to the square root of tornado's area has been improved substantially by the copula method. The obtained non-linear correlation coefficients vary with time and location. Long-Short Term Memory networks are used for the prediction of future property losses associated with tornadoes.
|
Authors who are presenting talks have a * after their name.
