Laser Scanning Systems in Highway and Safety Assessment

Anglický jazyk

Laser Scanning Systems in Highway and Safety Assessment

Na objednávku, dodanie 2-4 týždne

108.89 €

bežná cena: 120.99 €

O knihe

This book aims to promote the core understanding of a proper modelling of road traffic accidents by Neural Networks (NN) using traffic information and road geometry delineated from laser scanning data. The first two chapters of the book introduce the reader to laser-scanning technology with creative explanation and graphical illustrations, review and recent methods of extracting geometric road parameters. The third and fourth chapter present an optimizing support vector machine, ensemble trees and a novel method for networks. After introducing the principles of neural networks in chapter 5 and reviewing recent studies on accident modelling by neural networks in chapter 6, this book goes on exploring the details of NN and the performance in predicting the traffic accidents and their comparison with common data mining models. Chapter 7 presents deep learning applications in modelling accident data using feedforward, convolutional, recurrent neural. Chapter 8 develops an injury severity prediction model using feedforward NN. Chapter 9 improves the injury severity prediction model by replacing the traditional feedforward NN by a Convolutional Neural Network. In chapter 10, a Recurrent Neural Network (RNN), which accounts for temporal structure inherent in the accident data. Chapter 11 presents a procedure for modelling traffic accident with little data based on the concept of Transfer Learning. Chapter 12 presents the comparative study between NN models, support vector machine, and logistic regression for injury severity prediction. Finally, the methods of quantifying the importance of accident related factors in NN models are presented in chapter 13. This book aims to help graduate students, professionals, decision makers, and road planners in developing better traffic accident prediction models using advanced neural networks.