Papers by MAHDIEH ALLAHVIRANLOO
Transportmetrica B: Transport Dynamics, 2016
Chains of activities performed during the course of the day are interconnected such that particip... more Chains of activities performed during the course of the day are interconnected such that participation in one activity and the time allocated to that specific activity correspondingly influence the time-use behavior of a traveler along the course of the day. This points to the importance of analyzing trajectories of patterns as a set of activities with such specific characteristics as start time, duration, and sequence, rather than simply analyzing participation in each activity singularly. In this paper we present a methodology to answer a main question in the trajectory analysis: How to generate activity patterns trajectories, and how to conduct useful analysis that eventually makes inferences drawn from the time-use behavior of individuals applicable to the population-at-large possible? The methodology presented in this paper can be applied to synthetize chains of activities and their space-time distribution. It starts with clustering the activity patterns into a small set of representative patterns by using message passing algorithms, and then capturing the correlation among demographic profiles of travelers to the bundles of activities performed and their corresponding time sequence using multivariate probit models. We apply the methodology to two sets of data: (1) California household travel survey data for year 2000-2001, and (2) California household travel survey data for year 2010-2011. The longitudinal analysis performed in this work: (1) proves the robustness of proposed methodology in replicating time-use behavior and synthetizing activity chains, (2) reveals dynamics of changes in the trajectories of activity patterns during a 10-year time span, and (3) quantifies the influence of different socio-demographic variables on the trajectories of activities performed by travelers by implementing a statistical analysis on the distribution of estimates.
Transportation
GPS enabled devices, generating high-resolution spatial-temporal data, are opening new lines of p... more GPS enabled devices, generating high-resolution spatial-temporal data, are opening new lines of possibilities for transportation applications in both planning and research. Mining these rich and large datasets to infer the travel behavior of users and the activity patterns resulting from this behavior is the focus of this paper. Here we introduce a methodology that relies only on geocoded location data and socioeconomic characteristics to infer types of activities in which individuals engage at different locations in the network. At the census tract level, depending on the duration of the stop, arrival time and geographic distance to home location and previous activities, the type of activity is inferred using Adaptive Boosting algorithm. Then, using a model based on Markov Chains with Conditional Random Field to capture dependency between activity sequencing and individuals’ socioeconomic attributes, the spatial-temporal trajectory of activity/travel engagement is generated. The mo...
Transportation Research Part E: Logistics and Transportation Review, 2014
We argue that the selective vehicle routing problem is more appropriate than the conventional VRP... more We argue that the selective vehicle routing problem is more appropriate than the conventional VRP in handling uncertainty with limited resources. However, previous formulations of selective VRPs have all been deterministic. Three new formulations are proposed to account for different optimization strategies under uncertain demand (or utility) level: reliable, robust, and fuzzy selective vehicle routing problems. Three parallel genetic algorithms (PGAs) and a classic genetic algorithm are developed and compared to the deterministic solution. PGAs differ based on their communication strategies and diversity in sub-populations. Results show that a PGA, wherein communication between demes, or subpopulations, occurs in every generation and does not eliminate repeated chromosomes, outperforms other algorithms at the cost of higher computation time. A faster variation of PGA is used to solve the non-convex reliable selective VRP, robust selective VRP and the large-scale fuzzy selective VRP, consisting of 200 nodes. Large scale application demonstrates the value of fuzzy selective vehicle routing problem FSVRP in humanitarian logistics.
ABSTRACT The goal of this paper is to understand household priorities when the household members ... more ABSTRACT The goal of this paper is to understand household priorities when the household members are scheduling their activities during a given day. For this purpose, we reformulate the Household Activity Pattern Problem (HAPP), a network-based pickup and delivery problem with time windows, in the form of a goal-programming problem. Under this approach, and using the results of a previous analysis of travel diary data suggesting the existence of 8 relatively unique representative travel/activity patterns that individuals tend to perform, we set 6 different goals that measure the utility of performing an activity depending on the activity type and household characteristics. The target value for each of the goals is derived from the population of other households that belong to the same representative pattern, minimizing the pattern deviation among the households from the same group. We solve the proposed formulation under two scenarios: equal priorities for all goals in the objective function and calibrated priorities. For the calibration process we use a differential evolution search algorithm to efficiently explore the multiple priorities combinations. The results indicate that after calibration, the overall error measure, measured as the edit distances between the actual and predicted activity patterns, decreases by an average of 7.4%. Additionally, we observe that households do have different priorities depending on the representative pattern they are assigned to, validating the existence of the representative patterns and the relaxation of the original HAPP formulation.
ABSTRACT In order to capture the dependency between socio-demographic characteristics of individu... more ABSTRACT In order to capture the dependency between socio-demographic characteristics of individuals and their daily activities as well as the correlation among different activity types, we fit a multivariate probit (MVP) model to individual activity participation diary data. Due to the correlation among different activity types, analytical solutions under maximizing the likelihood function are intractable; in this paper, model parameters are estimated using parameter expansion and reparameterizaton with the Metropolis Hastings algorithm. The algorithm is based on Markov Chain Monte Carlo (MCMC) sampling techniques, with the assumption of normal and inverse Wishart distributions for regression parameters and unrestricted covariance matrix. Two sets of explanatory variables at individual and household levels are used in the study and the results indicate the influence of various socio-demographic variables on activity types that individuals select. A comparison between a multivariate probit model and an independent model, estimated on San Diego and Orange County travel data, indicates that higher accuracy is achieved using the correlated model in both in-sample and out-of sample data.
Studying travel behavior and activity engagement in an activity-based framework has been a focus ... more Studying travel behavior and activity engagement in an activity-based framework has been a focus of research for nearly half a century. A number of elegant and comprehensive models have been developed to address questions pertaining to activity participation, agenda formation, scheduling, and travel behavior of individuals. Despite the progress made in activity-based models, there is still a significant need for model improvements in the sense of modeling activity selection procedure and scheduling. In this paper, the authors propose a comprehensive model, which is the integration of discrete choice models, fuzzy concepts and Household Activity Pattern Problem (HAPP) to forecast household activity pattern based on socio-demographic characteristics. By using the values of probabilities obtained from a multivariate probit model applied to clustered households and mapping them to a set of fuzzy graphs, the authors compute the possibility of inclusion of an activity in the agenda. Activity scheduling and selection is then modeled as the outcome of a mixed integer optimization problem, in which the objective function is maximizing the expected desirability gained from activities and total saved time, subject to network connectivity, time windows, time budget and cost budget constraints.
Transportation Research Part B: Methodological, 2013
ABSTRACT The focus of this paper is to learn the daily activity engagement patterns of travelers ... more ABSTRACT The focus of this paper is to learn the daily activity engagement patterns of travelers using Support Vector Machines (SVMs), a modeling approach that is widely used in Artificial intelligence and Machine Learning. It is postulated that an individual’s choice of activities depends not only on socio-demographic characteristics but also on previous activities of individual on the same day. In the paper, Markov Chain models are used to study the sequential choice of activities. The dependencies among activity type, activity sequence and socio-demographic data are captured by employing hidden Markov models. In order to learn model parameters, we use sequential multinomial logit models (MNL) and multiclass Support Vector Machines (K-SVM) with two different dependency structures. In the first dependency structure, it is assumed that type of activity at time ‘t’ depends on the last previous activity and socio-demographic data, whereas in the second structure we assume that activity selection at time ‘t’ depends on all of the individual’s previous activity types on the same day and socio-demographic characteristics. The models are applied to data drawn from a set of California households and a comparison of the accuracy of estimation of activity types and their sequence in the agenda, indicates the superiority of K-SVM models over MNL. Additionally, we show that accuracy in estimating activity patterns increases using different sets of explanatory variables or tuning parameters of the kernel function in K-SVM.
Amirkabir Journal of Civil Engineering, Oct 23, 2019
Transportation Research Part B: Methodological
Transportation Research Part D: Transport and Environment
Sustainability
Traffic congestion is a major challenge in metropolitan areas due to economic and negative health... more Traffic congestion is a major challenge in metropolitan areas due to economic and negative health impacts. Several strategies have been tested all around the globe to relieve traffic congestion and minimize transportation externalities. Congestion pricing is among the most cited strategies with the potential to manage the travel demand. This study aims to investigate potential travel behavior changes in response to cordon pricing in Manhattan, New York. Several pricing schemes with variable cordon charging fees are designed and examined using an activity-based microsimulation travel demand model. The findings demonstrate a decreasing trend in the total number of trips interacting with the central business district (CBD) as the price goes up, except for intrazonal trips. We also analyze a set of other performance measures, such as Vehicle-Hours of Delay, Vehicle-Miles Traveled, and vehicle emissions. While the results show considerable growth in transit ridership (6%), single-occupan...
International Conference on Transportation and Development 2018
Transportation Research Part D: Transport and Environment
Transportation Research Part E: Logistics and Transportation Review
Transportation Research Record: Journal of the Transportation Research Board
Transportation Research Part C: Emerging Technologies
Journal of Transport & Health
Journal of Transport & Health
Uploads
Papers by MAHDIEH ALLAHVIRANLOO
In the model-building process, we initially cluster individuals in the sample based on their reported (one-day) activity patterns. Later, we argue and demonstrate that clustering activity/travel patterns in terms of such activity characteristics as type, duration, scheduling, and location can be an effective tool to capture preferential distributions of arrival time, departure time, and duration, which are unobservable inputs to activity-based travel models. Representative patterns are found based on two measures of dissimilarities between activity patterns, Sequence Alignment Method and Agenda dissimilarity, resulting in 8 clusters. A decision tree based on socio-demographics of individuals is fitted to infer the cluster to which each individual belongs.
Inference on agenda formation in each cluster is based on ensemble of three different modules—"multivariate probit model," "Markov chains with conditional random fields," and "adaptive boosting"—applied to individuals within each cluster. In each of these modules, the inputs are socio-demographic attributes of individuals, and the outputs are discrete outcomes indicating participation in each activity type.
Arrival time and activity duration inference for each activity type in each cluster, is performed using the adaptive boosting algorithm. Having identified the type of activities, and their arrival time and duration, activities are scheduled in the agenda using two approaches: decision rules, and Household Activity Pattern Problem (HAPP: a variation of pickup and delivery problem with time windows, (Recker, 1995)).
Testing the entire modeling system on an out-of-sample population—15% of the entire sample—shows that the model is able to predict on average 80.3% of daily activities of individuals; correct activities during 867 minutes of 1080 awake minutes in a day was predicted.