Year
2023
Season
Fall
Paper Type
Master's Thesis
College
College of Computing, Engineering & Construction
Degree Name
Master of Science in Civil Engineering (MSCE)
Department
Engineering
NACO controlled Corporate Body
University of North Florida. School of Engineering
First Advisor
Dr. Cigdem Akan
Second Advisor
Dr. Adel ElSafty
Third Advisor
Dr. Beyza Aslan
Department Chair
Alan Harris
Abstract
Extreme wind events, such as hurricanes and tropical storms, have devastating impacts on infrastructure, economy, and public safety in Florida. Traditional predictive methods, linear regression, logistic regression, and autoregressive integrated moving-average (ARIMA) models have limitations in accounting for non-stationarity and nonlinearity in meteorological data, leading to less accurate predictions. The primary objective of this study is to leverage the capabilities of deep learning algorithms, specifically, Feedforward Neural Networks (FFNNs), to develop a predictive model that outperforms traditional predictive methods in forecasting extreme wind events in Florida. The deep learning model was trained, validated, and tested using historical wind speed data obtained from 34 stations located throughout Florida. The performance of the model was evaluated and compared to the Support Vector Regression model (SVR) using measures such as Mean Square Error (MSE), Root Mean Square Error (RMSE), and Coefficient of Determination (R2). Results indicate that the feedforward neural network (FFNN) model achieved higher prediction accuracy than the Support Vector Regression (SVR) model. Deep learning offers a promising avenue for more accurate, adaptable, and comprehensive models for predicting extreme wind events in Florida.
Suggested Citation
Byemerwa, Irene Josephat, "Predicting Florida's extreme wind events: A deep learning approach" (2023). UNF Graduate Theses and Dissertations. 1232.
https://digitalcommons.unf.edu/etd/1232