Estimation of safety performance functions for urban intersections using various functional forms of the negative binomial regression model and a generalized Poisson regression model

Khattak, Muhammad Wisal; Pirdavani, Ali; Winne, Pieter De; Brijs, Tom; Backer, Hans De

doi:10.1016/j.aap.2020.105964

Cited by 32 publications

(30 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Besides that, other techniques such as the Bayesian network quantify occupational accident rates, fuzzy Bayesian networks for damaged equipment analysis, bow tie representation for occupational risk assessment, and Poisson models for occupational injury impacts modeling 4 . The negative binomial regression model is one of the most popular models because it simplifies estimation and performs better 79 . Cui et al 80 developed a Bayesian network model and a game‐theoretic model for risk assessment of third‐party damage to the pipelines.…”

Section: Application Of Machine Learning In Risk Analysismentioning

confidence: 99%

“…4 The negative binomial regression model is one of the most popular models because it simplifies estimation and performs better. 79 Cui et al 80 developed a Bayesian network model and a gametheoretic model for risk assessment of third-party damage to the pipelines. This work was done to overcome the probabilistic risk assessment in which human action and intentional act cannot model reliably.…”

Section: Application Of Machine Learning In Risk Analysismentioning

confidence: 99%

See 1 more Smart Citation

A review on supervised machine learning for accident risk analysis: Challenges in Malaysia

Chong

Razak

Biak

et al. 2022

Process Safety Progress

View full text Add to dashboard Cite

The new Fourth Industrial Revolution (IR 4.0) trend is driven by the concept of automation and artificial intelligence (AI). However, Malaysia is slightly behind Singapore in terms of adopting AI innovation among ASEAN countries. This paper aims to conduct a literature review of machine learning to overcome subjectivity and bias in risk ranking decision‐making. An introduction to machine learning concerning accident risk analysis is presented, and the challenges of its application in Malaysia are discussed. Existing machine learning features were evaluated to identify the feasible application in industrial accident analysis and ensure safety decision‐making consistency. This review observed how the IR 4.0 approaches were used in the risk analysis, especially on supervised machine learning. This study also highlights the finding from the previous works on challenges in utilizing supervised machine learning, which is the need to have publicly accessible large database from industries and agencies such as the Department of Occupational Safety and Health (DOSH) Malaysia for the development of algorithms, which can potentially improve accident risk analysis and safety, especially for Malaysian industries.

show abstract

Section: Application Of Machine Learning In Risk Analysismentioning

confidence: 99%

Section: Application Of Machine Learning In Risk Analysismentioning

confidence: 99%

A review on supervised machine learning for accident risk analysis: Challenges in Malaysia

Chong

Razak

Biak

et al. 2022

Process Safety Progress

View full text Add to dashboard Cite

show abstract

“…In road safety data, this restriction may be violated because the observed count may present a variance that is greater or smaller than expected, resulting in cases of overdispersion or underdispersion, respectively. The NB regression is often used to model RTAs, RAFs or RAIs with overdispersion, for example, the number of injury accidents on road bridges in Norway during 2010-2016 [18]; economic development and RAFs and RAIs with spatial panel    data analysis in Thailand during 2012-2016 [37]; the numbers of human deaths per crash in the Oromia region of Ethiopia [19] and safety performance functions for urban intersections of Antwerp in Belgium [20]. The CMP regression is an alternative model to the NB regression and its advantage of handling both overdispersed and underdispersed count.…”

Section: Model Selectionmentioning

confidence: 99%

“…Differences in the distribution of accidents, deaths, and injuries could be attributed to the variation between roads and their surroundings. Several studies estimate the frequency of RTAs and/or RAFs by traditional Poisson regression [17], Negative Binomial (NB) regression [18][19][20], and Conway-Maxwell Poisson (CMP) regression [21][22][23], as well as estimate the count data with excessive zeroes by zero-inflated regression models [24][25][26]. However, predictive models for deaths and/or injuries are limited in Thailand road safety.…”

Section: Introductionmentioning

confidence: 99%

Modelling Road Accident Fatalities with Underdispersion and Zero-inflated Counts

Simmachan

Wongsai

et al. 2022

Preprint

View full text Add to dashboard Cite

BackgroundThailand was rank second in the world in 2013 on the road accident fatality (RAF) rate, killing 36.2 of every 100,000 Thai peoples. In the past decade, during Songkran festival, the traditional Thai new year, the number of road traffic accidents (RTAs) was markedly higher than normal day life, but few studies have yet investigated this issue as the effect of festivity. The objective of this study was to investigate factors contributing to RAF using various count regression models.MethodsData of 20,229 accidents in 2015 were collected from the Department of Disaster Prevention and Mitigation, Thailand. Poisson, Conway–Maxwell–Poisson, and their Zero-Inflated versions were applied to analyze factors associated with the number of fatalities in an accident.ResultsThe RAFs in Thailand follow a count distribution with underdispersion and excessive zeros which is rare. The best fitting model, the ZICMP regression model returns significant predictors (road characteristics, weather conditions, environmental conditions, and month) on the number of fatalities in an accident. The model consists of the count part encapsulating both non-excess zeros and death counts and the zero-part representing the considerable number of zeros during the festival months. The estimated proportion of the zero-part is 0.275 accounting for 5,563 non-fatal accidents. More specifically, the excessive number of no deaths can be explained by the month factor. The mean number of fatalities was lower in the festive periods than other months, with the highest in November.ConclusionFor long, Thai authorities have put a lot of efforts and resources into improving road safety over the festival weeks, often they failed. This study indicates that people’s risk perception and public awareness of RAFs are mislead. Instead, nationwide road safety should have been announced by the authorities to raise the awareness of society towards everyday personal safety and the safety of others.

show abstract

“…Various studies have been conducted to develop SPFs at intersections (1,(19)(20)(21)(22)(23). Crash data have an overdispersion problem where the variance is greater than the average.…”

Section: Network Screeningmentioning

confidence: 99%

Development of New Performance Measures Based on Data Mining Weights for Hotspot Identification

Son

Park

Lee

et al. 2022

Transportation Research Record

View full text Add to dashboard Cite

In this study, new performance measures are proposed for hotspot identification in urban intersections. These measures reflect severity factor weights, which are determined based on data mining. To estimate the severity factor weights of crashes at urban intersections, the study utilizes tree-based random forest (RF) and extreme gradient boosting (XGB) methods. The importance of variables in the severity classification model is standardized and utilized for calculating the score of each crash, which is aggregated into intersections. The aggregated score is used as a dependent variable for the safety performance functions (SPFs) in the network screening process. To illustrate the under-dispersed severity score aggregation data, SPFs that follow the COM-Poisson distribution as well as the negative binomial (NB) are developed. Independent variables in SPFs set up intersection geometry elements that can be collected from online GIS services. Four additional performance measures are proposed, each reflecting a severity weight. Data about a total of 42,513 intersection crashes from 2017 to 2018 in South Korea were collected for crash injury severity analysis. Hotspot identification was performed on 81 intersections, and three consistency tests were conducted to validate the four measures. Tests show that the RF-based weighted [Formula: see text] and [Formula: see text] have the best consistency. Since the severity factor weights of each crash are reflected, intersections vulnerable to dangerous crashes can be analyzed in more detail. Using this method, it is expected that effective safety improvement project plans can be established with the input of safety managers in the future.

show abstract

Estimation of safety performance functions for urban intersections using various functional forms of the negative binomial regression model and a generalized Poisson regression model

Cited by 32 publications

References 51 publications

A review on supervised machine learning for accident risk analysis: Challenges in Malaysia

A review on supervised machine learning for accident risk analysis: Challenges in Malaysia

Modelling Road Accident Fatalities with Underdispersion and Zero-inflated Counts

Development of New Performance Measures Based on Data Mining Weights for Hotspot Identification

Contact Info

Product

Resources

About