1. UKM Learning and Research Repository
2. PHD Thesis / Tesis PHD
3. Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Title:	The analysis of travel mode choice using artificial neural networks and logistic regression
Authors:	Hussain Dhafir Hussain (P68229)
Supervisor:	Riza Atiq Abdullah O.K. Rahmat, Prof. Dato' Ir. Dr.
Keywords:	Transportation system Public transportation Prediction Universiti Kebangsaan Malaysia -- Dissertations
Issue Date:	11-Jun-2017
Description:	Baghdad, the capital city of Iraq is one of the largest cities in the Middle East in terms of population and land area; it is witnessing a significant number of changes and alterations in the transportation system. The city is highly congested partially caused by the fact that a large percentage of the population in the city prefers to use their private cars over any kind of public transportation. This contributed significantly to the ever worsening problem of traffic congestion as well as traffic accidents and air pollution. A proper solution is to encourage the commuters to use public means of transportation by focusing on the positive aspects of the public transport and the amelioration that comes with them leaving private cars usage for their daily trips. A survey was carried out in four known public areas with higher percentage of traffic congestion (north-east-west and south areas of Baghdad). Random sampling method was utilized in such manner that respondents chosen to symbolize the total demographic and socioeconomic characteristics, which include gender, age, income, trip types, and travelling behaviour. Logistic regression models were utilized for the purpose of prediction the probability of travel behaviour for all trips in the city. Logistic regression models helps to precise identification of the factors effecting people's transport choice as well as determining the policies that cannot be underestimated in encouraging travellers to ditch their private cars and start using public transportation means. The significance of cost, duration of the travel and waiting time as well as the parking cost in the city is discussed in this study. The results showed that the most suitable reduction in both cost and travel time is 45% and 50% respectively, on the other hand the results concerning the reduce waiting time showed that mode shift probabilities ranged from 52% probability of private car use with current reduce waiting time less than 10 minutes to 1% probability of private car use with reduce waiting time more than 30 minutes, other results ranging from 10% to 90%. As for parking cost, probability ranged between 96.3% in private cars with IQD 1000 in parking cost to 1% when the parking cost reached IQD 5000. Artificial Neural Network (ANN) models were used for the purpose of prediction of travelling behaviour for all trips in the city; ANN models were useful in inclusively identifying the factors affecting the travellers in choosing their mode of transportation. A Comparison between logistic regression models and Artificial Neural Network models was carried out to predict the transportation behaviour and detect the accuracy of predictability outcome. The final results showed that ANN models had the upper hand in predicting travel mode choice.,Ph.D.
Pages:	351
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

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