Automated Setting of Bus Schedule Coverage Using Unsupervised Machine Learning

Khiari, Jihed; Moreira-Matias, Luís; Cerqueira, Vítor; Cats, Oded

doi:10.1007/978-3-319-31753-3_44

Cited by 20 publications

(16 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traffic and commuter demand has also been considered in the work by Wang et al [27]. Other than employing optimization algorithms several deep learning techniques [22] have been applied in bus scheduling problems [15].…”

Section: Related Work and Challengesmentioning

confidence: 99%

Data-Driven Optimization of Public Transit Schedule

et al. 2019

View full text Add to dashboard Cite

Bus transit systems are the backbone of public transportation in the United States. An important indicator of the quality of service in such infrastructures is on-time performance at stops, with published transit schedules playing an integral role governing the level of success of the service. However there are relatively few optimization architectures leveraging stochastic search that focus on optimizing bus timetables with the objective of maximizing probability of bus arrivals at timepoints with delays within desired on-time ranges. In addition to this, there is a lack of substantial research considering monthly and seasonal variations of delay patterns integrated with such optimization strategies. To address these, this paper makes the following contributions to the corpus of studies on transit on-time performance optimization: (a) an unsupervised clustering mechanism is presented which groups months with similar seasonal delay patterns, (b) the problem is formulated as a single-objective optimization task and a greedy algorithm, a genetic algorithm (GA) as well as a particle swarm optimization (PSO) algorithm are employed to solve it, (c) a detailed discussion on empirical results comparing the algorithms are provided and sensitivity analysis on hyper-parameters of the heuristics are presented along with execution times, which will help practitioners looking at similar problems. The analyses conducted are insightful in the local context of improving public transit scheduling in the Nashville metro region as well as informative from a global perspective as an elaborate case study which builds upon the growing corpus of empirical studies using nature-inspired approaches to transit schedule optimization.

show abstract

Section: Related Work and Challengesmentioning

confidence: 99%

Data-Driven Optimization of Public Transit Schedule

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Moreira-Matias et al [23] introduced a generic and simple framework to empower the resistance of most induction learning algorithms and solve transportation problems. Khiari et al [24] proposed an unsupervised learning framework to determine the best schedule coverage to be assigned to a given PT network, and improve schedule reliability on a large scale.…”

Section: Related Workmentioning

confidence: 99%

Uncovering Distribution Patterns of High Performance Taxis from Big Trace Data

Tang

Sun

Kan

et al. 2017

IJGI

View full text Add to dashboard Cite

Abstract:The unbalanced distribution of taxi passengers in space and time affects taxi driver performance. Existing research has studied taxi driver performance by analyzing taxi driver strategies when the taxi is occupied. However, searching for passengers when vacant is costly for drivers, and it limits operational efficiency and income. Few researchers have taken the costs during vacant status into consideration when evaluating taxi driver performance. In this paper, we quantify taxi driver performance using the taxi's average efficiency. We propose the concept of a high-efficiency single taxi trip and then develop a quantification and evaluation model for taxi driver performance based on single trip efficiency. In a case study, we first divide taxi drivers into top drivers and ordinary drivers, according to their performance as calculated from their GPS traces over a week, and analyze the space-time distribution and operating patterns of the top drivers. Then, we compare the space-time distribution of top drivers to ordinary drivers. The results show that top drivers usually operate far away from downtown areas, and the distribution of top driver operations is highly correlated with traffic conditions. We compare the proposed performance-based method with three other approaches to taxi operation evaluation. The results demonstrate the accuracy and feasibility of the proposed method in evaluating taxi driver performance and ranking taxi drivers. This paper could provide empirical insights for improving taxi driver performance.

show abstract

“…This shortcoming is found in the reliability performance measurements, and in the analysis of possible unreliability causes and sources, as well as in the headway analysis reported in inferential statistics models. More precisely, first, in the extensive literature on the use of off-line AVL data, there are several studies in which anomalies are ignored, [26][27][28][29][30][31] even if there are exceptions in which data pruning, 32 days pruning 33 and data imputation methods 34 were performed. Second, Hammerle et al 35 mentioned AVL data anomalies in the calculation of headways, but they were not carefully analyzed and accurately measured, that is, the anomalies were intentionally neglected.…”

mentioning

confidence: 99%

“…TF and IOS were ignored in Feng and Figliozzi 37 in the calculation of the headway delay. Fourth, no previous studies [26][27][28][29][30][31][32][33][34][35][36][37] (i) distinguished between a TF and an IOS, which differently affect the passenger's viewpoint and (ii) rebuilt the original schedule at a stop to reflect the published time at the stop. Therefore, to the best of our knowledge, a gap exists in the literature on the accurate use of AVL raw data, both in the calculation of headway and of schedule deviation.…”

mentioning

confidence: 99%

Time Reliability Measures in bus Transport Services from the Accurate use of Automatic Vehicle Location raw Data

Barabino¹,

Francesco

Mozzoni³

2016

Quality & Reliability Eng

View full text Add to dashboard Cite

Archived automatic vehicle location (AVL) data are widely used in bus transportation, but typically contain anomalies, such as missing data points and unseen bus overtakings. These anomalies may alter the measurement of time reliability (i.e., headways and schedule time deviations) of buses at stops with respect to passenger experiences. However, in many studies, anomalies are ignored, neglected, or partially addressed. This paper investigates the effect of AVL anomalies on headways and schedule deviations, which are keys for regularity and punctuality evaluations and are well-practicized among bus operators worldwide. The analysis is carried out on six scenarios, each representing a combination of AVL data anomalies. The outcomes show that addressing anomalies result in more severe regularity measures, which point out criticalities suffered by passengers, whereas it can be skipped for the measurement of punctuality. In addition, an increase of 5% of missing data results in the increase of 5% in the coefficient of variation of the headway, but it does not lead to meaningful variations in punctuality measures. These results are useful for both practitioners interested in passenger-oriented measurement of time reliability and researchers working on the disclosure of the determinants of unreliability

show abstract

Automated Setting of Bus Schedule Coverage Using Unsupervised Machine Learning

Cited by 20 publications

References 11 publications

Data-Driven Optimization of Public Transit Schedule

Data-Driven Optimization of Public Transit Schedule

Uncovering Distribution Patterns of High Performance Taxis from Big Trace Data

Time Reliability Measures in bus Transport Services from the Accurate use of Automatic Vehicle Location raw Data

Contact Info

Product

Resources

About