Search

In this paper, the accelerated failure time (AFT) model is modified to analyze post-work grocery shopping activity duration. Much previous shopping duration analysis was conducted using the proportional hazard (PH) modeling approach. Once the proportionality assumption was violated, the traditional accelerated failure time (TAFT) model was usually selected as an alternative modeling approach. However, a TAFT model only has covariates with non-proportional and time-dependent effects on the hazard overtime while a PH model only accommodates covariates with proportional and time-independent effects. Neither of them considers the possibility that some of covariates may have proportional and time-independent effects and some may have non-proportional and time-dependent effects on the hazard value in one model. To address this issue, the paper generalizes the TAFT model and develops a modified accelerated failure time (MAFT) model to accommodate both time-dependent and time-independent covariates for activity duration analysis. Checking on the proportionality assumption indicates that the assumption is not valid in the post-work grocery shopping activity data extracted from the 2017 National Household Travel Survey (NHTS) conducted by the U.S. Department of Transportation (USDOT). Both TAFT and MAFT models are developed for comparisons and analysis. The empirical and statistical results show that there do exist two different types of covariates affecting shopping activity duration, including covariates only with proportional and time-independent effects (i.e. working duration, commute travel time) and those with non-proportional and time-dependent effects. The MAFT model can capture the subtleties in various types of covariate effects and help better understand how those covariates affect activity duration overtime. This paper also shows the importance to develop a flexible duration model with both time-dependent and time-independent covariates for accurately evaluating travel demand management (TDM) policies, like flexible work hours. [ABSTRACT FROM AUTHOR]

With the convergence of various broadcasting systems, the amount of content available in mobile terminals including IPTV has significantly increased. In this paper, we propose a system that enables users to schedule programs considering both section video clips and full video clips based on the user detection method with similar preference. And, since the system constituting the contents can be classified according to the program, the proposed method can store a program desired by the user, and thus create and schedule a kind of individual channel. Experimental results show that the proposed method has a higher prediction accuracy; this is accomplished by comparing existing channel recommendation methods with the program recommendation methods proposed in this paper. [ABSTRACT FROM AUTHOR]

In this paper we introduce a novel algorithm, called Exogenous Fitness, to calculate the Fitness of subnational entities and we apply it to the states of Brazil. In the last decade, several indices were introduced to measure the competitiveness of countries by looking at the complexity of their export basket. Tacchella et al (2012) developed a non-monetary metric called Fitness. In this paper, after an overview about Brazil as a whole and the comparison with the other BRIC countries, we introduce a new methodology based on the Fitness algorithm, called Exogenous Fitness. Combining the results with the Gross Domestic Product per capita (GDPp), we look at the dynamics of the Brazilian states in the Fitness-Income plane. Two regimes are distinguishable: one with high predictability and the other with low predictability, showing a deep analogy with the heterogeneous dynamics of the World countries. Furthermore, we compare the ranking of the Brazilian states according to the Exogenous Fitness with the ranking obtained through two other techniques, namely Endogenous Fitness and Economic Complexity Index. [ABSTRACT FROM AUTHOR]

Adapting methods from complex system analysis, this paper analyzes the features of the complex relationship between wage inequality and the development and industrialization of a country. Development is understood as a combination of a monetary index, GDP per capita, and a recently introduced measure of a country’s economic complexity: Fitness. Initially the paper looks at wage inequality on a global scale, over the time period 1990–2008. Our empirical results show that globally the movement of wage inequality along with the ongoing industrialization of countries has followed a longitudinally persistent pattern comparable to the one theorized by Kuznets in the fifties: countries with an average level of development suffer the highest levels of wage inequality. Next, the study narrows its focus on wage inequality within the United States. By using data on wages and employment in the approximately 3100 US counties over the time interval 1990–2014, it generalizes the Fitness-Complexity metric for geographic units and industrial sectors, and then investigates wage inequality between NAICS industries. The empirical time and scale dependencies are consistent with a relation between wage inequality and development driven by institutional factors comparing countries, and by change in the structural compositions of sectors in a homogeneous institutional environment, such as the counties of the United States. [ABSTRACT FROM AUTHOR]

The ability to produce timely and accurate flu forecasts in the United States can significantly impact public health. Augmenting forecasts with internet data has shown promise for improving forecast accuracy and timeliness in controlled settings, but results in practice are less convincing, as models augmented with internet data have not consistently outperformed models without internet data. In this paper, we perform a controlled experiment, taking into account data backfill, to improve clarity on the benefits and limitations of augmenting an already good flu forecasting model with internet-based nowcasts. Our results show that a good flu forecasting model can benefit from the augmentation of internet-based nowcasts in practice for all considered public health-relevant forecasting targets. The degree of forecast improvement due to nowcasting, however, is uneven across forecasting targets, with short-term forecasting targets seeing the largest improvements and seasonal targets such as the peak timing and intensity seeing relatively marginal improvements. The uneven forecasting improvements across targets hold even when “perfect” nowcasts are used. These findings suggest that further improvements to flu forecasting, particularly seasonal targets, will need to derive from other, non-nowcasting approaches. [ABSTRACT FROM AUTHOR]

This paper examines social class-based differences in influence in online review contexts. We explore four mechanisms for how a review writer’s social class may affect readers’ evaluations of the organization. First, we argue that, via a “contagion” process, organizations reviewed by higher-class individuals will be evaluated more positively than organizations reviewed by lower-class individuals. Second, we expect that higher-class reviewers will be seen as more knowledgeable; thus, their opinions will be more influential in shaping others’ opinions. Third, we expect that reviewers will be seen more influential when they review organizations that match their social class. Fourth, we expect people to be more influenced by those who share their own class background. A large-scale observational study of reviews (N = 1,234,665) from finds support for the contagion, the organization-reviewer social class matching, and the reviewer-participant social matching hypotheses, but disconfirms the hypothesis that higher-class reviewers are always treated as having more expertise. Two experimental studies (N = 354 and N = 638) demonstrate that reviewer class plays a causal role in both a contagion process and in an assumption of higher-class knowledge process, but do not provide evidence for the reviewer-participant social matching hypothesis. [ABSTRACT FROM AUTHOR]

PubMed is a free search engine for biomedical literature accessed by millions of users from around the world each day. With the rapid growth of biomedical literature—about two articles are added every minute on average—finding and retrieving the most relevant papers for a given query is increasingly challenging. We present Best Match, a new relevance search algorithm for PubMed that leverages the intelligence of our users and cutting-edge machine-learning technology as an alternative to the traditional date sort order. The Best Match algorithm is trained with past user searches with dozens of relevance-ranking signals (factors), the most important being the past usage of an article, publication date, relevance score, and type of article. This new algorithm demonstrates state-of-the-art retrieval performance in benchmarking experiments as well as an improved user experience in real-world testing (over 20% increase in user click-through rate). Since its deployment in June 2017, we have observed a significant increase (60%) in PubMed searches with relevance sort order: it now assists millions of PubMed searches each week. In this work, we hope to increase the awareness and transparency of this new relevance sort option for PubMed users, enabling them to retrieve information more effectively. [ABSTRACT FROM AUTHOR]

Interactions between countries originate from diverse aspects such as geographic proximity, trade, socio-cultural habits, language, religions, etc. Geopolitics studies the influence of a country’s geographic space on its political power and its relationships with other countries. This work reveals the potential of Wikipedia mining for geopolitical study. Actually, Wikipedia offers solid knowledge and strong correlations among countries by linking web pages together for different types of information (e.g. economical, historical, political, and many others). The major finding of this paper is to show that meaningful results on the influence of country ties can be extracted from the hyperlinked structure of Wikipedia. We leverage a novel stochastic matrix representation of Markov chains of complex directed networks called the reduced Google matrix theory. For a selected small size set of nodes, the reduced Google matrix concentrates direct and indirect links of the million-node sized Wikipedia network into a small Perron-Frobenius matrix keeping the PageRank probabilities of the global Wikipedia network. We perform a novel sensitivity analysis that leverages this reduced Google matrix to characterize the influence of relationships between countries from the global network. We apply this analysis to two chosen sets of countries (i.e. the set of 27 European Union countries and a set of 40 top worldwide countries). We show that with our sensitivity analysis we can exhibit easily very meaningful information on geopolitics from five different Wikipedia editions (English, Arabic, Russian, French and German). [ABSTRACT FROM AUTHOR]

Few if any natural resource systems are completely understood and fully observed. Instead, there almost always is uncertainty about the way a system works and its status at any given time, which can limit effective management. A natural approach to uncertainty is to allocate time and effort to the collection of additional data, on the reasonable assumption that more information will facilitate better understanding and lead to better management. But the collection of more data, either through observation or investigation, requires time and effort that often can be put to other conservation activities. An important question is whether the use of limited resources to improve understanding is justified by the resulting potential for improved management. In this paper we address directly a change in value from new information collected through investigation. We frame the value of information in terms of learning through the management process itself, as well as learning through investigations that are external to the management process but add to our base of understanding. We provide a conceptual framework and metrics for this issue, and illustrate them with examples involving Florida scrub-jays (Aphelocoma coerulescens). [ABSTRACT FROM AUTHOR]

This paper develops a semi-nonparametric Poisson regression model to analyze motor vehicle crash frequency data collected from rural multilane highway segments in California, US. Motor vehicle crash frequency on rural highway is a topic of interest in the area of transportation safety due to higher driving speeds and the resultant severity level. Unlike the traditional Negative Binomial (NB) model, the semi-nonparametric Poisson regression model can accommodate an unobserved heterogeneity following a highly flexible semi-nonparametric (SNP) distribution. Simulation experiments are conducted to demonstrate that the SNP distribution can well mimic a large family of distributions, including normal distributions, log-gamma distributions, bimodal and trimodal distributions. Empirical estimation results show that such flexibility offered by the SNP distribution can greatly improve model precision and the overall goodness-of-fit. The semi-nonparametric distribution can provide a better understanding of crash data structure through its ability to capture potential multimodality in the distribution of unobserved heterogeneity. When estimated coefficients in empirical models are compared, SNP and NB models are found to have a substantially different coefficient for the dummy variable indicating the lane width. The SNP model with better statistical performance suggests that the NB model overestimates the effect of lane width on crash frequency reduction by 83.1%. [ABSTRACT FROM AUTHOR]

In environmental epidemiology, it is often encountered that multiple time series data with a long-term trend, including seasonality, cannot be fully adjusted by the observed covariates. The long-term trend is difficult to separate from abnormal short-term signals of interest. This paper addresses how to estimate the long-term trend in order to recover short-term signals. Our case study demonstrates that the current spline smoothing methods can result in significant positive and negative cross-correlations from the same dataset, depending on how the smoothing parameters are chosen. To circumvent this dilemma, three classes of time series smoothers are proposed to detrend time series data. These smoothers do not require fine tuning of parameters and can be applied to recover short-term signals. The properties of these smoothers are shown with both a case study using a factorial design and a simulation study using datasets generated from the original dataset. General guidelines are provided on how to discover short-term signals from time series with a long-term trend. The benefit of this research is that a problem is identified and characteristics of possible solutions are determined. [ABSTRACT FROM AUTHOR]

Accurate and fast path computation is essential for applications such as onboard navigation systems and traffic network routing. While a number of heuristic algorithms have been developed in the past few years for faster path queries, the accuracy of them are always far below satisfying. In this paper, we first develop an agglomerative graph partitioning method for generating high balanced traverse distance partitions, and we constitute a three-level graph model based on the graph partition scheme for structuring the urban road network. Then, we propose a new hierarchical path computation algorithm, which benefits from the hierarchical graph model and utilizes a region pruning strategy to significantly reduce the search space without compromising the accuracy. Finally, we present a detailed experimental evaluation on the real urban road network of New York City, and the experimental results demonstrate the effectiveness of the proposed approach to generate optimal fast paths and to facilitate real-time routing applications. [ABSTRACT FROM AUTHOR]

Objectives: As gentrification continues in New York City as well as other urban areas, residents of lower socioeconomic status maybe at higher risk for residential displacement. Yet, there have been few quantitative assessments of the health impacts of displacement. The objective of this paper is to assess the association between displacement and healthcare access and mental health among the original residents of gentrifying neighborhoods in New York City. Methods: We used 2 data sources: 1) 2005–2014 American Community Surveys to identify gentrifying neighborhoods in New York City, and 2) 2006–2014 Statewide Planning and Research Cooperative System. Our cohort included 12,882 residents of gentrifying neighborhoods in 2006 who had records of emergency department visits or hospitalization at least once every 2 years in 2006–2014. Rates of emergency department visits and hospitalizations post-baseline were compared between residents who were displaced and those who remained. Results: During 2006–2014, 23% were displaced. Compared with those who remained, displaced residents were more likely to make emergency department visits and experience hospitalizations, mainly due to mental health (Rate Ratio = 1.8, 95% confidence interval = 1.5, 2.2), after controlling for baseline demographics, health status, healthcare utilization, residential movement, and the neighborhood of residence in 2006. Conclusions: These findings suggest negative impacts of displacement on healthcare access and mental health, particularly among adults living in urban areas and with a history of frequent emergency department visits or hospitalizations. [ABSTRACT FROM AUTHOR]

Projecting the long-term trends in energy demand is an increasingly complex endeavor due to the uncertain emerging changes in factors such as climate and policy. The existing energy-economy paradigms used to characterize the long-term trends in the energy sector do not adequately account for climate variability and change. In this paper, we propose a multi-paradigm framework for estimating the climate sensitivity of end-use energy demand that can easily be integrated with the existing energy-economy models. To illustrate the applicability of our proposed framework, we used the energy demand and climate data in the state of Indiana to train a Bayesian predictive model. We then leveraged the end-use demand trends as well as downscaled future climate scenarios to generate probabilistic estimates of the future end-use demand for space cooling, space heating and water heating, at the individual household and building level, in the residential and commercial sectors. Our results indicated that the residential load is much more sensitive to climate variability and change than the commercial load. Moreover, since the largest fraction of the residential energy demand in Indiana is attributed to heating, future warming scenarios could lead to reduced end-use demand due to lower space heating and water heating needs. In the commercial sector, the overall energy demand is expected to increase under the future warming scenarios. This is because the increased cooling load during hotter summer months will likely outpace the reduced heating load during the more temperate winter months. [ABSTRACT FROM AUTHOR]

The social functions of urbanized areas are highly dependent on and supported by the convenient access to public transportation systems, particularly for the less privileged populations who have restrained auto ownership. To accurately evaluate the public transit accessibility, it is critical to capture the spatiotemporal variation of transit services. This can be achieved by measuring the shortest paths or minimum travel time between origin-destination (OD) pairs at each time-of-day (e.g. every minute). In recent years, General Transit Feed Specification (GTFS) data has been gaining popularity for between-station travel time estimation due to its interoperability in spatiotemporal analytics. Many software packages, such as ArcGIS, have developed toolbox to enable the travel time estimation with GTFS. They perform reasonably well in calculating travel time between OD pairs for a specific time-of-day (e.g. 8:00 AM), yet can become computational inefficient and unpractical with the increase of data dimensions (e.g. all times-of-day and large network). In this paper, we introduce a new algorithm that is computationally elegant and mathematically efficient to address this issue. An open-source toolbox written in C++ is developed to implement the algorithm. We implemented the algorithm on City of St. George’s transit network to showcase the accessibility analysis enabled by the toolbox. The experimental evidence shows significant reduction on computational time. The proposed algorithm and toolbox presented is easily transferable to other transit networks to allow transit agencies and researchers perform high resolution transit performance analysis. [ABSTRACT FROM AUTHOR]

Staff education is considered key to quality of early childhood education and care (ECEC) programs. However, findings about associations between staff education and children’s outcomes have been inconsistent. We conducted a systematic review and meta-analysis of associations between ECEC staff education and child outcomes. Searches of Medline, PsycINFO, and ERIC, websites of large datasets and reference sections of all retrieved articles were conducted. Eligible studies provided a statistical link between staff education and child outcomes for preschool-aged children in ECEC programs. Titles, abstracts and paper reviews as well as all data extraction were conducted by two independent raters. Of the 823 studies reviewed for eligibility, 39 met our inclusion criteria. Research in this area is observational in nature and subject to the inherent biases of that research design. Results from our systematic review were hampered by heterogeneity in how staff education was defined, variability in whose education was measured and the child outcomes that were assessed. However, overall the qualitative summary indicates that associations between staff education and childhood outcomes are non-existent to very borderline positive. In our meta-analysis of more homogeneous studies we identified certain positive, albeit very weak, associations between staff education and children’s language outcomes (specifically, vocabulary and letter word identification) and no significant association with a mathematics outcome (WJ Applied Problems). Thus, our findings suggest that within the range of education levels found in the existing literature, education is not a key driver of child outcomes. However, since we only explored levels of education that were reported in the literature, our findings cannot be used to argue for lowering education standards in ECEC settings. [ABSTRACT FROM AUTHOR]

Adaptive management involves learning-oriented decision making in the presence of uncertainty about the responses of a resource system to management. It is implemented through an iterative sequence of decision making, monitoring and assessment of system responses, and incorporating what is learned into future decision making. Decision making at each point is informed by a value or objective function, for example total harvest anticipated over some time frame. The value function expresses the value associated with decisions, and it is influenced by system status as updated through monitoring. Often, decision making follows shortly after a monitoring event. However, it is certainly possible for the cadence of decision making to differ from that of monitoring. In this paper we consider different combinations of annual and biennial decision making, along with annual and biennial monitoring. With biennial decision making decisions are changed only every other year; with biennial monitoring field data are collected only every other year. Different cadences of decision making combine with annual and biennial monitoring to define 4 scenarios. Under each scenario we describe optimal valuations for active and passive adaptive decision making. We highlight patterns in valuation among scenarios, depending on the occurrence of monitoring and decision making events. Differences between years are tied to the fact that every other year a new decision can be made no matter what the scenario, and state information is available to inform that decision. In the subsequent year, however, in 3 of the 4 scenarios either a decision is repeated or monitoring does not occur (or both). There are substantive differences in optimal values among the scenarios, as well as the optimal policies producing those values. Especially noteworthy is the influence of monitoring cadence on valuation in some years. We highlight patterns in policy and valuation among the scenarios, and discuss management implications and extensions. [ABSTRACT FROM AUTHOR]

This paper forecasts the 2016 canine Anaplasma spp. seroprevalence in the United States from eight climate, geographic and societal factors. The forecast’s construction and an assessment of its performance are described. The forecast is based on a spatial-temporal conditional autoregressive model fitted to over 11 million Anaplasma spp. seroprevalence test results for dogs conducted in the 48 contiguous United States during 2011–2015. The forecast uses county-level data on eight predictive factors, including annual temperature, precipitation, relative humidity, county elevation, forestation coverage, surface water coverage, population density and median household income. Non-static factors are extrapolated into the forthcoming year with various statistical methods. The fitted model and factor extrapolations are used to estimate next year’s regional prevalence. The correlation between the observed and model-estimated county-by-county Anaplasma spp. seroprevalence for the five-year period 2011–2015 is 0.902, demonstrating reasonable model accuracy. The weighted correlation (accounting for different sample sizes) between 2015 observed and forecasted county-by-county Anaplasma spp. seroprevalence is 0.987, exhibiting that the proposed approach can be used to accurately forecast Anaplasma spp. seroprevalence. The forecast presented herein can a priori alert veterinarians to areas expected to see Anaplasma spp. seroprevalence beyond the accepted endemic range. The proposed methods may prove useful for forecasting other diseases. [ABSTRACT FROM AUTHOR]

Overweight and obesity are highly prevalent in the population of the United States, affecting roughly 2/3 of Americans. These diseases, along with their associated conditions, are a major burden on the healthcare industry in terms of both dollars spent and effort expended. Volitional weight loss is attempted by many, but weight regain is common. The ability to predict which patients will lose weight and successfully maintain the loss versus those prone to regain weight would help ease this burden by allowing clinicians the ability to skip treatments likely to be ineffective. In this paper we introduce a new windowed approach to the persistent homology signal processing algorithm that, when paired with a modified, semimetric version of the Hausdorff distance, can differentiate the two groups where other commonly used methods fail. The novel approach is tested on accelerometer data gathered from an ongoing study at the University of Michigan. While most standard approaches to signal processing show no difference between the two groups, windowed persistent homology and the modified Hausdorff semimetric show a clear separation. This has significant implications for clinical decision making and patient care. [ABSTRACT FROM AUTHOR]

In peer-to-peer networks, application relays have been commonly used to provide various networking services. The service performance often improves significantly if a relay is selected appropriately based on its network location. In this paper, we studied the location-aware relay discovery and selection problem for large-scale P2P streaming networks. In these large-scale and dynamic overlays, it incurs significant communication and computation cost to discover a sufficiently large relay candidate set and further to select one relay with good performance. The network location can be measured directly or indirectly with the tradeoffs between timeliness, overhead and accuracy. Based on a measurement study and the associated error analysis, we demonstrate that indirect measurements, such as King and Internet Coordinate Systems (ICS), can only achieve a coarse estimation of peers’ network location and those methods based on pure indirect measurements cannot lead to a good relay selection. We also demonstrate that there exists significant error amplification of the commonly used “best-out-of-K” selection methodology using three RTT data sets publicly available. We propose a two-phase approach to achieve efficient relay discovery and accurate relay selection. Indirect measurements are used to narrow down a small number of high-quality relay candidates and the final relay selection is refined based on direct probing. This two-phase approach enjoys an efficient implementation using the Distributed-Hash-Table (DHT). When the DHT is constructed, the node keys carry the location information and they are generated scalably using indirect measurements, such as the ICS coordinates. The relay discovery is achieved efficiently utilizing the DHT-based search. We evaluated various aspects of this DHT-based approach, including the DHT indexing procedure, key generation under peer churn and message costs. [ABSTRACT FROM AUTHOR]

This paper studies a reliable facility location problem with facility protection that aims to hedge against random facility disruptions by both strategically protecting some facilities and using backup facilities for the demands. An Integer Programming model is proposed for this problem, in which the failure probabilities of facilities are site-specific. A solution approach combining Lagrangian Relaxation and local search is proposed and is demonstrated to be both effective and efficient based on computational experiments on random numerical examples with 49, 88, 150 and 263 nodes in the network. A real case study for a 100-city network in Hunan province, China, is presented, based on which the properties of the model are discussed and some managerial insights are analyzed. [ABSTRACT FROM AUTHOR]

Traffic state estimation from the floating car system is a challenging problem. The low penetration rate and random distribution make available floating car samples usually cover part space and time points of the road networks. To obtain a wide range of traffic state from the floating car system, many methods have been proposed to estimate the traffic state for the uncovered links. However, these methods cannot provide traffic state of the entire road networks. In this paper, the traffic state estimation is transformed to solve a missing data imputation problem, and the tensor completion framework is proposed to estimate missing traffic state. A tensor is constructed to model traffic state in which observed entries are directly derived from floating car system and unobserved traffic states are modeled as missing entries of constructed tensor. The constructed traffic state tensor can represent spatial and temporal correlations of traffic data and encode the multi-way properties of traffic state. The advantage of the proposed approach is that it can fully mine and utilize the multi-dimensional inherent correlations of traffic state. We tested the proposed approach on a well calibrated simulation network. Experimental results demonstrated that the proposed approach yield reliable traffic state estimation from very sparse floating car data, particularly when dealing with the floating car penetration rate is below 1%. [ABSTRACT FROM AUTHOR]

HIV incidence estimates are used to monitor HIV-1 infection in the United States. Use of laboratory biomarkers that distinguish recent from longstanding infection to quantify HIV incidence rely on having accurate knowledge of the average time that individuals spend in a transient state of recent infection between seroconversion and reaching a specified biomarker cutoff value. This paper describes five estimation procedures from two general statistical approaches, a survival time approach and an approach that fits binomial models of the probability of being classified as recently infected, as a function of time since seroconversion. We compare these procedures for estimating the mean duration of recent infection (MDRI) for two biomarkers used by the U.S. National HIV Surveillance System for determination of HIV incidence, the Aware BED EIA HIV-1 incidence test (BED) and the avidity-based, modified Bio-Rad HIV-1/HIV-2 plus O ELISA (BRAI) assay. Collectively, 953 specimens from 220 HIV-1 subtype B seroconverters, taken from 5 cohorts, were tested with a biomarker assay. Estimates of MDRI using the non-parametric survival approach were 198.4 days (SD 13.0) for BED and 239.6 days (SD 13.9) for BRAI using cutoff values of 0.8 normalized optical density and 30%, respectively. The probability of remaining in the recent state as a function of time since seroconversion, based upon this revised statistical approach, can be applied in the calculation of annual incidence in the United States. [ABSTRACT FROM AUTHOR]

Rainy weather conditions could result in significantly negative impacts on driving on freeways. However, due to lack of enough historical data and monitoring facilities, many regions are not able to establish reliable risk assessment models to identify such impacts. Given the situation, this paper provides an alternative solution where the procedure of risk assessment is developed based on drivers’ subjective questionnaire and its performance is validated by using actual crash data. First, an ordered logit model was developed, based on questionnaire data collected from Freeway G15 in China, to estimate the relationship between drivers’ perceived risk and factors, including vehicle type, rain intensity, traffic volume, and location. Then, weighted driving risk for different conditions was obtained by the model, and further divided into four levels of early warning (specified by colors) using a rank order cluster analysis. After that, a risk matrix was established to determine which warning color should be disseminated to drivers, given a specific condition. Finally, to validate the proposed procedure, actual crash data from Freeway G15 were compared with the safety prediction based on the risk matrix. The results show that the risk matrix obtained in the study is able to predict driving risk consistent with actual safety implications, under rainy weather conditions. [ABSTRACT FROM AUTHOR]