The impact of real-time information on bus ridership in New York City

In the past few years, numerous mobile applications have made it possible for public transit passengers to find routes and/or learn about the expected arrival time of their transit vehicles. Though these services are widely used, their impact on overall transit ridership remains unclear. The objective of this research is to assess the effect of real-time information provided via web-enabled and mobile devices on public transit ridership. An empirical evaluation is conducted for New York City, which is the setting of a natural experiment in which a real-time bus tracking system was gradually launched on a borough-by-borough basis beginning in 2011. Panel regression techniques are used to evaluate bus ridership over a three year period, while controlling for changes in transit service, fares, local socioeconomic conditions, weather, and other factors. A fixed effects model of average weekday unlinked bus trips per month reveals an increase of approximately 118 trips per route per weekday (median increase of 1.7% of weekday route-level ridership) attributable to providing real-time information. Further refinement of the fixed effects model suggests that this ridership increase may only be occurring on larger routes; specifically, the largest quartile of routes defined by revenue miles of service realized approximately 340 additional trips per route per weekday (median increase of 2.3% per route). Although the increase in weekday route-level ridership may appear modest, on aggregate these increases exert a substantial positive effect on farebox revenue. The implications of this research are critical to decision-makers at the country’s transit operators who face pressure to increase ridership under limited budgets, particularly as they seek to prioritize investments in infrastructure, service offerings, and new technologies.     

Real time location prediction with taxi-GPS data streams

The prediction of the destination location at the time of pickup is an important problem with potential for substantial impact on the efficiency of a GPS-enabled taxi service. While this problem has been explored earlier in the batch data set-up, we propose in this paper new solutions in the streaming data set-up. We examine four incremental learning methods using a damped window model namely, Multivariate multiple regression, Spherical-spherical regression, Randomized spherical K-NN regression and an Ensemble of these methods for their effectiveness in solving the destination prediction problem. The performance of these methods on several large datasets are evaluated using suitably chosen metrics and they were also compared with some other existing methods. We found that the Multivariate multiple regression method has the best performance in terms of prediction accuracy but the Spherical-spherical regression method is the best performer when we take into account the accuracy time trade-off criterion. The next pickup location problem, where we are interested in predicting the next pickup location for a taxi given the dropoff location coordinates of the previous trip as input is also considered and the aforementioned methods are examined for their suitability using real world datasets. As in the case of destination prediction problem, here also we find that the Multivariate multiple regression method gives better performance than the rest when we consider prediction accuracy but the Spherical-spherical regression method is the best performer when the accuracy-time trade-off criterion is taken into account.     

Utilizing naturalistic driving data for in-depth analysis of driver lane-keeping behavior in rain: Non-parametric MARS and parametric logistic regression modeling approaches

It is known that adverse weather conditions can affect driver performance due to reduction in visibility and slippery surface conditions. Lane keeping is one of the main factors that might be affected by weather conditions. Most of the previous studies on lane keeping have investigated driver lane-keeping performance from driver inattention perspective. In addition, the majority of previous lane-keeping studies have been conducted in controlled environments such as driving simulators. Therefore, there is a lack of studies that investigate driver lane-keeping ability considering adverse weather conditions in naturalistic settings. In this study, the relationship between weather conditions and driver lane-keeping performance was investigated using the SHRP2 naturalistic driving data for 141 drivers between 19 and 89 years of age. Moreover, a threshold was introduced to differentiate lane keeping and lane changing in naturalistic driving data. Two lane-keeping models were developed using the logistic regression and multivariate adaptive regression splines (MARS) to better understand factors affecting driver lane-keeping ability considering adverse weather conditions. The results revealed that heavy rain can significantly increase the standard deviation of lane position (SDLP), which is a very widely used method for analyzing lane-keeping ability. It was also found that traffic conditions, driver age and experience, and posted speed limits have significant effects on driver lane-keeping ability. An interesting finding of this study is that drivers have a better lane-keeping ability in roadways with higher posted speed limits. The results from this study might provide better insights into understanding the complex effect of adverse weather conditions on driver behavior.     

Have Emission Control Areas (ECAs) harmed port efficiency in Europe?

Ports in the European Union and North America have enforced environmental regulations on controlling SO_x and NOx emissions from ships in their coastal areas known as Emission Control Areas (ECAs). This study uses two-stage approaches to examine whether ECA regulations impact the efficiency of ports operating in such areas. First, port efficiencies are estimated using non-radial slacks-based measure (SBM) Data Envelopment Analysis (DEA) models. The efficiency scores estimated by the SBM DEA models are then regressed on explanatory variables, including the ECA factor, and macroeconomic indicators using bootstrapped truncated regression (BTR) models. Panel data is collected on countries in EU ECAs and non-ECAs regarding such input variables as capital and labor, with cargo as an output variable. The results indicate that ECA regulations can harm port efficiency, reflecting concerns of policy-makers and industrial managers: the average efficiency loss from an ECA designation amounts to 0.058–0.066 on a scale of 0–1, accounting for a 15–18% loss from ECA ports’ average efficiency scores.     

Reliable route guidance: A case study from Chicago

Reliable route guidance can be obtained by solving the reliable a priori shortest path problem, which finds paths that maximize the probability of arriving on time. The goal of this paper is to demonstrate the benefits and applicability of such route guidance using a case study. An adaptive discretization scheme is first proposed to improve the efficiency in computing convolution, a time-consuming step used in the reliable routing algorithm to obtain path travel time distributions. Methods to construct link travel time distributions from real data in the case study are then discussed. Particularly, the travel time distributions on arterial streets are estimated from linear regression models calibrated from expressway data. Numerical experiments demonstrate that optimal paths are substantially affected by the reliability requirement in rush hours, and that reliable route guidance could generate up to 5-15% of travel time savings. The study also verifies that existing algorithms can solve large-scale problems within a reasonable amount of time.     

支持向量回归在轮轨Hertz弹性接触计算中的应用

空间轮轨接触是铁道车辆系统动力学仿真对象模型中的重要柔性环节，基于材料线弹性假设的Hertz理论，可以给出两弹性体一般接触的复杂的隐式非线性关系.支持向量回归是统计学习理论中提出的新一代学习算法．轮轨接触计算可以使用它离线学习到一个显式的非线性关系．其逼近精度和表示复杂度可得到很好的折中．有利于铁道车辆系统动力学仿真的在线计算，     

一种基于Logistic回归和分类树的客户信用评估方法研究

探讨了基于Logistic回归和分类树的客户信用评估。从众多客户信用度影响因子中选择一些探索性变量，建立了相应的关联性测试模型。描述了Logistic回归和分类树的具体算法以及相关概念，如期望信息、信息增益等，然后分别使用Logistic回归模型和分类树对客户信用进行了测试评估。比较分析结果表明，分类树模型具有较低的错误分类率和较好的灵活性，但对计算资源的要求较高，且很大程度上依赖于观测数据。     

中国青年人动脉血pH值与地理因素的关系

目的建立中国青年人(18-39岁)动脉血血气分析血pH值正常参考值与地理因素之间的回归模型。方法收集全国各地区数据及各地区地理因素指标,对中国青年人动脉血pH值与地理因素的关系进行研究。结果得出中国青年人动脉血pH值与地理因素之间的回归方程。结论如果知道中国任何地区的地理因素指标,就可以应用此方程方便地推算出中国青年人动脉血pH值的正常参考值,实际应用于各地临床血气分析pH正常值参考值的预测。     

基于GIS的Logistic回归在区域滑坡空间预测中的应用

巫山新址西区作为试验区,选取地形地貌、岩性特征、地质构造等区域斜坡稳定性影响因子,以MAPGIS地理信息系统为技术支撑,将全区按10m×10m方格划分为14450个单元,每个格网单元作为一个样本,建立Logistic回归模型,进行区域滑坡空间预测。结果表明:以滑坡发生概率0.163为判据,试验区内发生滑坡地段的判对率为72.08%,不发生滑坡地段的判对率为81.44%。根据计算结果可以得到区域斜坡稳定性分区以及进一步的地质灾害风险评价。     

Constrained Run-to-Run Optimization for Batch Process Based on Support Vector Regression Model

Introduction Batch and semi-batch processes are very im-portant in the fine chemicals industry to manufac-ture the low-volume and high-value products suchas biochemicals, crystals and speciality chemicals.In the control of batch process, the major objectiveis to improve the end-of-batch product properties.When the control objective is to optimize a perfor-mance index at the end of batch cycle, the problemis called the end-point optimization problem. Tosolve the problem,the traditional approach…