A household car ownership modeling is crucial in understanding the impact on an individual's or a family's travel behavior in traveling demand analysis. Trips or tours as a unit of analysis can be used in the modeling of car ownership demand for analyzing travel needs. Machine learning is widely used to describe a car owner's decision since the machine learning model was specifically designed to give more accurate predictions through a variety of mechanisms. This research presents car ownership modeling using two types of machine learning models, including decision trees and neural networks. The impacts of socio-demographic attributes on household car ownership demand are discussed and compared against these two models after adding the main attributes of variables from tour-based models. Data was collected from 2,015 households surveyed in Khon Kaen Province, Thailand, conducted in 2015. The outcomes indicate that the machine learning model can be used to predict household car ownership. It also found that when using the default parameters across all datasets, whereas the neural networks provide a more accurate result than the decision tree algorithm. However, in cases where the household car ownership prediction from the dataset with add attributes of the key variable used in tour-based models. In that case, the neural networks algorithm would give a prediction accuracy that corresponded to results as found from the prediction using a dataset with only the household's socio-demographic attributes.

Because the numbers of cars reflect each person's travel behaviors for each specific location, the car ownership demand model plays a dominant role in analysis of the travel demand in order to understand each area's individual and household travel behaviors. However, the study project for the master plan of the Khon Kaen expressway represented imbalanced data; namely, the majority class and the minority class were not equal. Before developing a machine learning model, this study suggested a solution to balance the data by using oversampling and under-sampling techniques. The data, which had been improved with SMOTE (Synthetic Minority Oversampling Technique) and kNN (k-nearest neighbors) (k = 5), demonstrated a better effect than the other algorithms that were studied. The TPR (true positive rate) for the rural and suburban areas, which are types of regions with very different imbalance ratios, was calculated before balancing the data at 46.9 % and 46.4 %. As a result, the TPR values were 63.5 % and 54.4 %, respectively, following the data balancing.

In terms of the travel demand prediction from the household car ownership model, if the imbalanced data were used to support the transportation policy via a machine learning model, it would negatively affect the algorithm training process. The data on household car ownership obtained from the study project for the expressway preparation in the Khon Kaen Province (2015) was an unbalanced dataset. In other words, the number of members of the minority class is lower than the rest of the answer classes. The result is a bias in data classification. Consequently, this research suggested balancing the datasets with cost-sensitive learning methods, including decision trees, k-nearest neighbors (kNN), and naive Bayes algorithms. Before creating the 3-class model, a k-folds cross-validation method was applied to classify the datasets to define true positive rate (TPR) for the model’s performance validation. The outcome indicated that the kNN algorithm demonstrated the best performance for the minority class data prediction compared to other algorithms. It provides TPR for rural and suburban area types, which are region types with very different imbalance ratios, before balancing the data of 46.9% and 46.4%. After balancing the data (MCN1), TPR values were 84.4% and 81.4%, respectively.

In developing countries, motorcycle riders normally attempt to stop at their desired locations during queue formation on signalized intersection approaches. Under mixed-traffic conditions, motorcycle positioning in a queue affects the operational and safety performance of the intersection. This study aimed to identify factors influencing motorcycle riders’ stopping locations at signalized urban intersections. This study applied Unmanned Aerial Vehicles (UAVs) to observe the stopping behavior of 1413 motorcycle riders on 24 approaches from 10 signalized intersections in Thailand (N = 1413). Multinomial logistic regression analysis was used to determine the relationship between the stopping locations of motorcycle riders and rider- and motorcycle-related variables and traffic- and environmental-related variables. The statistical analyses presented a Cox and Snell R2 and Nagelkerke R2 of 0.466 and 0.499, respectively, indicating that the model accounted for almost 50% of the variation among the five stopping locations of motorcycle riders. The results showed that, under mixed-traffic conditions in Thailand with left-hand traffic, motorcycle riders intending to turn right, the morning peak period, the presence of shadows, motorcycle riders not wearing helmets, the presence of a larger vehicle in the queue, and the density of desired stopping locations significantly influenced the motorcyclists’ choice of stopping locations on signalized intersection approaches. Practical policy-related recommendations drawn from the findings are provided to improve motorcyclists’ safety on signalized intersection approaches.

The objective of this research was to evaluate the efficiency of the motorcycle-specific pausing zone under the Bike Box Project (BBP). This zone is designated at the very front of the waiting line for the traffic light. The indicators used to measure the efficiency comprised the lost time when starting up at the beginning of the green light, the change of the proportion of the motorcycle pausing points, the accident statistics, and the worthiness for investment. Two intersections in urban areas were used for the analyses and evaluation, which were carried out to compare between the period prior to and after the project began. The study indicated that after the project began, the start-up lost time could be reduced by roughly 31-46% during peak hours, with approximately 73-78% of motorcycles pausing in the green zone (Bike Box Zone). The number and severity of the traffic accidents after 1-year of BBP implementation significantly decreased. The results of the study showed the efficiency of traffic management, safety, and worthiness for investment. It is thus appropriate to extend the project to other areas in the country and abroad, especially in urban areas where motorcycles are greatly used.