東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Development of a Deep Learning-Based...

Fordjour, George Kwabena,

Linked to FindBook

Google Book

Amazon

博客來

Development of a Deep Learning-Based Flood Forecasting Tool With Firefly Algorithm for the Blackburn Fork Watershed, Tennessee /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Development of a Deep Learning-Based Flood Forecasting Tool With Firefly Algorithm for the Blackburn Fork Watershed, Tennessee // George Kwabena Fordjour.
Author:	Fordjour, George Kwabena,
Description:	1 electronic resource (99 pages)
Notes:	Source: Masters Abstracts International, Volume: 85-11.
Contained By:	Masters Abstracts International85-11.
Subject:	Hydrologic sciences. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31148679
ISBN:	9798382593326

Development of a Deep Learning-Based Flood Forecasting Tool With Firefly Algorithm for the Blackburn Fork Watershed, Tennessee /
Fordjour, George Kwabena,

Development of a Deep Learning-Based Flood Forecasting Tool With Firefly Algorithm for the Blackburn Fork Watershed, Tennessee /George Kwabena Fordjour. - 1 electronic resource (99 pages)

Source: Masters Abstracts International, Volume: 85-11.

Throughout history, floods have been the most frequent natural disaster, impacting communities worldwide. They account for a staggering 40% of all such events. In the United States, the devastation wrought by flooding has been particularly severe. Developing robust flood forecasting tools is crucial for mitigating these impacts. While traditional physical models have long played a role, Machine Learning (ML) and Deep Learning (DL) techniques are increasingly augmenting these methods. This study presents a flood forecasting tool that leverages DL models optimized with a meta-heuristic search algorithm called the Firefly Algorithm (FFA). The tool forecasts water levels in the Blackburn Fork River watershed, Tennessee, at 4-, 5-, and 6-hour lead times. Three specific deep learning architectures were employed: Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and One-dimensional Convolutional Neural Networks (1D CNN). To enhance the accuracy of these models, the FFA technique was used for optimization. The resulting hybrid models (FFA-LSTM, FFA-GRU, FFA-1D CNN) were then compared to their standalone counterparts (SA-LSTM, SA-GRU, SA-1D CNN), each optimized using a heuristic random search algorithm. The comparison was conducted across two flood scenarios (Scenario #1 and Scenario #2) at the various lead times. The hybrid models demonstrated promising results, particularly in reducing bias as evidenced by improved percent bias (PBIAS) and Kling-Gupta efficiency (KGE) metrics. While the improvement in root mean squared error (RMSE) and mean absolute error (MAE) was modest in some cases, the hybrid models showed a slight advantage. However, all models struggled with predicting very high-water levels, especially at longer lead times. Among the hybrid models, FFA-GRU emerged as the most effective, consistently performing well across various evaluation metrics. This flood forecasting tool, utilizing optimized deep learning models, offers a significant advancement for issuing timely flood warnings and facilitating evacuations during flood events. Future research could focus on further refining the models to improve accuracy, particularly for extreme water level forecasts.

English

ISBN: 9798382593326Subjects--Topical Terms:

3168407
Hydrologic sciences.
Subjects--Index Terms:

Machine Learning

Development of a Deep Learning-Based Flood Forecasting Tool With Firefly Algorithm for the Blackburn Fork Watershed, Tennessee /
LDR:03688nmm a22004333i 4500 001 2438794
005 20260727135833.5
006 m o d
007 cr|nu||||||||
008 261011s2024 miu||||||m |||||||eng d
020 $a 9798382593326
035 $a (MiAaPQD)AAI31148679
035 $a AAI31148679
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Fordjour, George Kwabena, $e author. $3 3830203
245 1 0 $a Development of a Deep Learning-Based Flood Forecasting Tool With Firefly Algorithm for the Blackburn Fork Watershed, Tennessee / $c George Kwabena Fordjour.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (99 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 85-11.
500 $a Advisors: Kalyanapu, Alfred J. Committee members: Talbert, Doug; Datta, Tania.
502 $b M.S. $c Tennessee Technological University $d 2024.
520 $a Throughout history, floods have been the most frequent natural disaster, impacting communities worldwide. They account for a staggering 40% of all such events. In the United States, the devastation wrought by flooding has been particularly severe. Developing robust flood forecasting tools is crucial for mitigating these impacts. While traditional physical models have long played a role, Machine Learning (ML) and Deep Learning (DL) techniques are increasingly augmenting these methods. This study presents a flood forecasting tool that leverages DL models optimized with a meta-heuristic search algorithm called the Firefly Algorithm (FFA). The tool forecasts water levels in the Blackburn Fork River watershed, Tennessee, at 4-, 5-, and 6-hour lead times. Three specific deep learning architectures were employed: Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and One-dimensional Convolutional Neural Networks (1D CNN). To enhance the accuracy of these models, the FFA technique was used for optimization. The resulting hybrid models (FFA-LSTM, FFA-GRU, FFA-1D CNN) were then compared to their standalone counterparts (SA-LSTM, SA-GRU, SA-1D CNN), each optimized using a heuristic random search algorithm. The comparison was conducted across two flood scenarios (Scenario #1 and Scenario #2) at the various lead times. The hybrid models demonstrated promising results, particularly in reducing bias as evidenced by improved percent bias (PBIAS) and Kling-Gupta efficiency (KGE) metrics. While the improvement in root mean squared error (RMSE) and mean absolute error (MAE) was modest in some cases, the hybrid models showed a slight advantage. However, all models struggled with predicting very high-water levels, especially at longer lead times. Among the hybrid models, FFA-GRU emerged as the most effective, consistently performing well across various evaluation metrics. This flood forecasting tool, utilizing optimized deep learning models, offers a significant advancement for issuing timely flood warnings and facilitating evacuations during flood events. Future research could focus on further refining the models to improve accuracy, particularly for extreme water level forecasts.
546 $a English
590 $a School code: 0390
650 4 $a Hydrologic sciences. $3 3168407
650 4 $a Environmental engineering. $3 548583
653 $a Machine Learning
653 $a Root mean squared error
653 $a Mean absolute error
653 $a Kling-Gupta efficiency
690 $a 0543
690 $a 0775
690 $a 0388
690 $a 0800
710 2 $a Tennessee Technological University. $b Civil Engineering. $e degree granting institution. $3 3820554
720 1 $a Kalyanapu, Alfred J. $e degree supervisor.
773 0 $t Masters Abstracts International $g 85-11.
790 $a 0390
791 $a M.S.
792 $a 2024
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31148679