published journal article
published journal article
Phd Dissertation
This thesis is concerned with specifying and estimating multivariate models in discrete data settings. The models are applied to several empirical applications with an emphasis in banking and monetary history. The approaches presented here are of central importance in model evaluation, policy analysis, and prediction. The first chapter develops a framework for estimating multivariate treatment effect models in the presence of sample selection. The methodology deals with several important issues prevalent in program evaluation, including non-random treatment assignment, endogeneity, and discrete outcomes. The framework is applied to evaluate the effectiveness of bank recapitalization programs and their ability to resuscitate the financial system. This paper presents a novel bank-level data set and employs the new methodology to jointly model a bank’s decision to apply for assistance, the central bank’s decision to approve or decline the assistance, and the bank’s performance. The article offers practical estimation tools to unveil new answers to important regulatory and government intervention questions. The second chapter examines an important but often overlooked obstacle in multivariate discrete data models which is the proper specification of endogenous covariates. Endogeneity can be modeled as latent or observed, representing competing hypotheses about the outcomes of interest. This paper highlights the use of existing Bayesian model comparison techniques to understand the nature of endogeneity. Consideration of both observed and latent modeling approaches is emphasized in two empirical applications. The first application examines linkages for banking contagion and the second application evaluates the impact of education on socioeconomic outcomes. The third chapter, which is joint work with Professor Ivan Jeliazkov, studies the formulation of the likelihood function for simultaneous equation models for discrete data. The approach rests on casting the required distribution as the invariant distribution of a suitably defined Markov chain. The derivation resolves puzzling paradoxes highlighted in earlier work, shows that such models are theoretically coherent, and offers simple and intuitive linkages to the better understood analysis of continuous outcomes. The new methodology is employed in two applications involving simultaneous equation models of (i) female labor supply and family financial stability, and (ii) the interactions between health and wealth.
research report
Parking infrastructure is suffering from congestion as the number of vehicles circulating in urban areas is growing and expansion is not a cost-effective solution. In parallel, developments in autonomous vehicle technology mean that driverless vehicles are predicted to be in circulation by the 2020s and makeup 40% of vehicle travel by the 2040s. Expected benefits of autonomous vehicle travel include reduced congestion through vehicle sharing and reduced walking distance for passengers who can be dropped off chauffeur-style by autonomous vehicles. However, empty vehicle cruising, or the case in which autonomous vehicles cannot efficiently locate parking and circle instead, can potentially increase congestion. Given that this new technology has the potential to exacerbate existing congestion issues, it is necessary to develop a solution for parking congestion integrated with autonomous vehicles. Our project addresses this issue by providing a full-stack solution including sensors to monitor occupancy, Fog systems to perform local data pre- processing, and SDR radios to communicate with autonomous vehicles.
published journal article
Vehicular ad hoc networks (VANETs) offer a promising way to prevent accidents, facilitate eco-friendly driving, and provide more accurate real-time traffic information. This article describes the three different communication pathways incorporated by VANETs and briefly outlines potential applications. While there are still communication problems to solve within these complex systems, concerns about privacy, liability, and security are the chief obstacles that prevent progress towards large-scale implementation.
conference paper
In this paper we present some of our research findings derived from a series of research activities funded by the California PATH program to commemorate the occasion of the establishment of the PATH program 20 years ago. The major theme woven by these research efforts is the development of more effective traffic management tools that help tame unbridled traffic congestion in California, and the major contributions include a better understanding of traveler behavior, improved methods for obtaining origin-destination demand matrices, and increased modeling and control capabilities
research report
This report presents the results of Phase I of a multi-year research effort on “Field Investigation of Advanced Vehicle Reidentification Techniques and Detector Technologies,” and extends previous PATH research by the authors on MOU 336 “Section-Related Measures of Traffic System Performance: Prototype Field Implementation.” The focus of this research included the following: significant expansion and enhancement of the ILD-based vehicle reidentification system at a major signalized intersection in Irvine, California to address reidentification of turning vehicles in addition to through vehicles; derivation of improved estimates of fundamental real-time traffic parameters such as speed, volume and vehicle class from single loop detector inductive signatures; development of a new technique for on-line real-time intersection level of service estimation; implementation of a capability for communicating real-time traffic performance data to operators in the City of Irvine Transportation Management Center (TMC); development of a prototype real-time web-site for internet-based access to performance data from the study intersection in Irvine (and other sites in the future); initial testing of a new state-of-the-art detector card (the IST-222, from IST, Inc.); and an initial study of video image processing for future detector data fusion of video and loop signature data. The very encouraging results obtained to date for signalized intersection application of the vehicle reidentification approach suggest that further development and improvement of the vehicle reidentification algorithms for this application would clearly be of value. Keywords vehicle signature, inductive loop detector, single loop speed estimation, vehicle classification, vehicle reidentification, signalized intersection, level of service, detector card, data fusion, web-site
book/book chapter
Phd Dissertation
This dissertation presents the vehicle reidentification problem formulated as a lexicographic optimization problem. The lexicographic optimization formulation is a preemptive multi-objective formulation that combines goal programming, classification, and Bayesian analysis techniques. The details of field implementation and data collection design are also presented. The solution of the vehicle reidentification problem has the potential to yield reliable section measures such as travel times and densities, and enables the measurement of specific dynamic origin/destination demands as well as the development of new algorithms for ATMIS (Advanced Transportation Management and Information Systems) implementations of the approach using conventional surveillance infrastructure. Freeway inductive loop data from SR-24 in Lafayette, California, demonstrates that robust results can be obtained under different traffic flow conditions. A discussion is also presented of the application of section densities in a dynamic origin/destination demand estimation framework as an example of the usefulness of this approach. The use of existing surveillance infrastructure coupled with this approach could allow development of widespread applications in Intelligent Transportation Systems (ITS).
