Author(s)

N. O. Nawari

Abstract

It's been more than five years since a major storm hit a major US city, however hurricane researchers have estimated that the next one could cause as much as $150 billion worth of damage. An understanding of the damage mechanisms due to a variety of natural hazards such as wind, storm surges, and tsunamis windstorm hazard is integral to determining the best design and construction practices. Therefore, an anatomical analysis of a large suite of damaged structures in similar extreme events, such as Florida coastal areas, promises to reveal causes of failure in coastal construction and how best to prevent similar damage in the future. Prediction and classification of hurricane buildings damage involves many sources of uncertain data that make it a difficult task using conventional prediction models. This work seeks to elucidate the application of data mining algorithms in the prediction and classification of damage due to hurricane and tornadoes forces. The research focuses on the conceptual and applied frameworks for the data mining models to assist in the prediction, assessment, and classification of buildings damages caused by server windstorms. Keywords: wind storms, building damages, prediction, data mining. 1 Introduction High winds, airborne projectiles, wind-driven water, sea surges, and flooding are among the hazards that threaten buildings and their occupancies. The record for US property insurance tells the story. Annual claims for wind damage claims tally hundreds of millions of dollars. Florida homeowners pay some of the nation's highest insurance premiums; in a recent poll, despite a housing crisis, an

Keywords

wind storms, building damages, prediction, data mining