Please use this identifier to cite or link to this item: https://ktisis.cut.ac.cy/handle/10488/18394
Title: Maximising Accuracy and Efficiency of Traffic Accident Prediction Combining Information Mining with Computational Intelligence Approaches and Decision Trees
Authors: Tambouratzis, Tatiana 
Souliou, Dora 
Chalikias, Miltiadis 
Gregoriades, Andreas 
Major Field of Science: Engineering and Technology
Field Category: Electrical Engineering - Electronic Engineering - Information Engineering
Issue Date: 30-Dec-2014
Source: Journal of Artificial Intelligence and Soft Computing Research,2014, vol. 4 no. 1 pp. 31-42
Volume: 4
Issue: 1
Start page: 31
End page: 42
Journal: Journal of Artificial Intelligence and Soft Computing Research 
Abstract: The development of universal methodologies for the accurate, efficient, and timely prediction of traffic accident location and severity constitutes a crucial endeavour. In this piece of research, the best combinations of salient accident-related parameters and accurate accident severity prediction models are determined for the 2005 accident dataset brought together by the Republic of Cyprus Police. The optimal methodology involves: (a) information mining in the form of feature selection of the accident parameters that maximise prediction accuracy (implemented via scatter search), followed by feature extraction (implemented via principal component analysis) and selection of the minimal number of components that contain the salient information of the original parameters, which combined bring about an overall 74.42% reduction in the dataset dimensionality; (b)accidentseveritypredictionviaprobabilisticneuralnetworksandrandomforests,both of which independently accomplish over 96% correct prediction and a balanced proportionofunder-andover-estimationsofaccidentseverity. Anexplanationofthesuperiority of the optimal combinations of parameters and models is given, as is a comparison with existing accident classification/prediction approaches.
ISSN: 2449-6499
DOI: 10.2478/jaiscr-2014-0023
Rights: © Sciendo
Type: Article
Affiliation : European University Cyprus 
University of Piraeus 
National Technical University of Athens 
Appears in Collections:Άρθρα/Articles

CORE Recommender
Show full item record

Page view(s)

28
Last Week
1
Last month
4
checked on Nov 28, 2020

Google ScholarTM

Check

Altmetric


Items in KTISIS are protected by copyright, with all rights reserved, unless otherwise indicated.