Keemin Sohn

Dynamic Origin–Destination Flow Estimation Using Cellular Communication System

This paper aims to develop a new approach for the estimation of dynamic origin-destination (O-D) flow using cell phones as traffic probes. The state-space model, which depends on the autoregressive dynamics of O-D flow and the time series for link volume counts, was adopted. Unlike a direct approach that uses sample O-D flows extracted from the cell-based location data as additional observations, an indirect approach is proposed wherein the assignment map in the model is derived from the passing time at observation locations and the path choice proportion. A probe phones passing time at a certain point in a cell was approximated with its entry and exit times at cell boundaries. The average path choice proportion was also estimated using cell-based trajectories of probe phones. The simulation experiments confirmed that the approach was successfully applicable to the real-world freeway network. The results suggest that the O-D flows estimated from the present approach are promising in that the mean absolute error ratio was smaller than the case wherein only historical O-D flows are concerned. A sensitivity analysis revealed that the present approach met the requirements for an urban area encompassing a huge number of cell phone users in a microcell system.

Space-Based Passing Time Estimation on a Freeway Using Cell Phones as Traffic Probes

This paper examines the usability of mobile cellular networks to obtain traffic information on a freeway. The question of whether a mobile station (cell phone) can play an acceptable role as a probe for collecting traffic information on a freeway is examined. A space-based approach, wherein the probe vehicles transmit information to roadside devices as they pass through reference points, is exploited rather than a time-based approach, wherein the probe vehicles report information for every specific instant of time. The latter has been of concern to most researchers interested in the use of a mobile cellular network for collecting traffic data. First, a simple analytical model is introduced to address the usability of cell phones as traffic probes and to pinpoint which factors affect the qualification of probe phones when the space-based approach is adopted. Second, simulation experiments are also employed to deal with more realistic traffic conditions as supplementary tools for the analytical model. Finally, the actual traffic data on a freeway was considered to validate the above two hypothesized traffic conditions. The findings show that there are three main factors that affect the qualification of cell phones as a traffic probe: (1) the speed profile of the probe phone in cell coverage; (2) the variability of handoff location where the probe phone switches its jurisdictional cell; and (3) the locational relationship between a reference point and a speed jump (or drop) point in cell coverage.

Identifying the Impact on Land Prices of Replacing At-grade or Elevated Railways with Underground Subways in the Seoul Metropolitan Area

The Seoul metropolitan government (SMG) is considering the replacement of at-grade or elevated railways penetrating the city’s central area with underground subways. The railways once played a key role in forming main transit corridors in the early development stages of the city of Seoul. Now, however, they are a burden in the efforts of the SMG to enhance livability in the vicinity of urban railways. Over the past few decades, the at-grade elevated railways have led to economic deterioration by severing neighbourhoods and causing serious environmental problems such as noise, vibration and an unattractive landscape. The present study focused on laying the empirical groundwork for the SMG to replace at-grade or elevated railways with underground subways. An aggregate-level regression model, in a form similar to a hedonic price model, was developed to identify the net influence of the existence of at-grade or elevated railways on nearby land values. Potential variables that would account for the price of land along urban railways were chosen based on insights and empirical results from previous studies. Statistical tests have verified a significant difference in several variables according to whether a station area belongs to at-grade (or elevated) railways or underground subways. As a result of the regression model, it was confirmed that the land price of areas along at-grade or elevated railways are much less than those along underground railways, all else being equal.

Classification of crystal structure using a convolutional neural network

A deep-machine-learning technique based on a convolutional neural network (CNN) is introduced. It has been employed for the classification of crystal system, extinction group and space group for given powder X-ray diffraction patterns of inorganic materials.

Transportation Deficiencies for Older Adults in Seoul, South Korea

South Korea expects a nearly tenfold increase between 2000 and 2018 in its share of adults older than 65 years of age. This population is likely to experience deficiencies in its transportation options as its physical abilities deteriorate, which will necessitate new policies to provide mobility for older adults. This study investigated deficiencies in transportation for a representative sample of 812 individuals, 65 years or older, surveyed in 2011 in Seoul, South Korea. The survey contained 160 questions about individual, household, transportation, and built environment characteristics. The respondents that occasionally or frequently could not engage in desired out-of-home activities because of a lack of transportation options were defined as transportation deficient, compared with those that indicated that they never or rarely lacked transportation. The probability of transportation deficiency decreased for older adults who knew how reach destinations on public transportation, lived within walking distance of the subway, lived in the neighborhood for more than 20 years, and lived within walking distance of places where they could meet other older adults. The probability of transportation deficiency increased for people who were 75 or older, had a physical disability, were male and had given up driving, had a low income, lived with children, and lived in areas with difficult conditions for pedestrians. The policy recommendations highlighted an improved pedestrian environment to facilitate walking, public transit improvements, and information systems that would make bus transit in particular a more viable option for older adults. Policies were needed to take into account the varied health and physical abilities of older adults.

Systematic Control of Experimental Inconsistency in Combinatorial Materials Science

We developed a method to systematically control experimental inconsistency, which is one of the most troublesome and difficult problems in high-throughput combinatorial experiments. The topic of experimental inconsistency is never addressed, even though all scientists in the field of combinatorial materials science face this very serious problem. Experimental inconsistency and material property were selected as dual objective functions that were simultaneously optimized. Specifically, in an attempt to search for promising phosphors with high reproducibility, photoluminescence (PL) intensity was maximized, and experimental inconsistency was minimized by employing a multiobjective evolutionary optimization-assisted combinatorial materials search (MOEO combinatorial material search) strategy. A tetravalent manganese-doped alkali earth germanium/titanium oxide system was used as a model system to be screened using MOEO combinatorial materials search. As a result of MOEO reiteration, we identified a halide-detached deep red phosphor with improved PL intensity and reliable reproducibility.

An extremely simple macroscale electronic skin realized by deep machine learning

Complicated structures consisting of multi-layers with a multi-modal array of device components, i.e., so-called patterned multi-layers, and their corresponding circuit designs for signal readout and addressing are used to achieve a macroscale electronic skin (e-skin). In contrast to this common approach, we realized an extremely simple macroscale e-skin only by employing a single-layered piezoresistive MWCNT-PDMS composite film with neither nano-, micro-, nor macro-patterns. It is the deep machine learning that made it possible to let such a simple bulky material play the role of a smart sensory device. A deep neural network (DNN) enabled us to process electrical resistance change induced by applied pressure and thereby to instantaneously evaluate the pressure level and the exact position under pressure. The great potential of this revolutionary concept for the attainment of pressure-distribution sensing on a macroscale area could expand its use to not only e-skin applications but to other high-end applications such as touch panels, portable flexible keyboard, sign language interpreting globes, safety diagnosis of social infrastructures, and the diagnosis of motility and peristalsis disorders in the gastrointestinal tract.

Optimizing Train-Stop Positions Along a Platform to Distribute the Passenger Load More Evenly Across Individual Cars

Crowding in metro trains is a major factor in determining both the passenger service level and the operator supply level. An uneven distribution in passenger load across individual cars of a train exacerbates the overall capacity loading of a metro transit system. A loading diversity factor has been adopted to adjust the effect when computing the capacity of a metro train. The passenger preference for a specific car of a train was found to depend upon minimizing the walking distance at destination stations. This paper is focused on the possibility that a passenger load could be more evenly dispersed by varying train-stop positions. This paper proposes a mathematical programming model to find the optimal train-stop position at each station of a hypothesized metro line. The objective function is set to minimize the discrepancies in passenger loading across individual cars. After applying a genetic algorithm to solve the proposed model, differentiating train-stop positions considerably improved the distribution of passenger loading.

An Expectation-Maximization Algorithm to Estimate the Integrated Choice and Latent Variable Model

As computing capability has grown dramatically, the transport choice model has rigorously included latent variables. However, integrated latent and choice variable (ICLV) models are hampered by a serious problem that is caused by the maximum simulated likelihood method. The method cannot properly reproduce the true coefficients, which is a problem that is often referred to as a lack of empirical identification. In particular, the problem is exacerbated particularly when an ICLV model is calibrated based on cross-sectional data. An expectation-maximization (EM) algorithm has been successfully employed to calibrate a random coefficient choice model, but it has never been applied to the calibration of an ICLV model. In this study, an EM algorithm was adapted to calibrate an ICLV model, and it successfully reproduced the true coefficients in the model. The main contribution of adopting an EM algorithm was to simplify the calibration procedure by decomposing the procedure into three well known econometric prob...

Deep-Learning Architectures to Forecast Bus Ridership at the Stop and Stop-To-Stop Levels for Dense and Crowded Bus Networks

ABSTRACT The conventional transit assignment models that depend on either probabilistic or deterministic theory have failed to accurately estimate rider demand for dense and crowded bus transit networks. It is well known that the existing models are so blunt that they cannot accommodate the impact of miscellaneous changes in activity and transportation systems on bus demand. Recently, artificial neural networks (ANNs) have been refocused after two monumental breakthroughs: Big-data and a novel pre-training method. A deep-learning model, which simply represents an ANN with multiple hidden layers, has had a great success in recognizing images, human voices, and handwritten texts. The present study adopted a deep-learning model to forecast bus ridership at the stop and stop-to-stop levels. While the stop-level model, which had insufficient training data, suffered from an overfitting of the data, the stop-to-stop-level model showed good performance both in training and testing. The success of the latter model is owed to a larger sample size compared with the former model. This represents the first meaningful attempt to apply a data-driven approach to forecasting transportation demand.