Machine learning approach for risk-based inspection screening assessment

Rachman, Andika; Ratnayake, R. M. Chandima

doi:10.1016/j.ress.2019.02.008

Cited by 83 publications

(36 citation statements)

References 82 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This makes many characteristics at the component level or system level unable to be modeled, such as interdependence, a common cause of failure, which yields a decline in modeling capability. According to the criterion of computation cost, our approach had small computational cost in contrast to some approaches that use machine learning [41]. For the criterion of maturity, our approach was less mature than Bayesian probability based methods, which have been widely applied in practice with many publications.…”

Section: Resultsmentioning

confidence: 99%

A Spatial Statistic Based Risk Assessment Approach to Prioritize the Pipeline Inspection of the Pipeline Network

Hou

Dong

2020

Energies

View full text Add to dashboard Cite

The identification of high risk regions is an important aim of risk-based inspections (RBIs) in pipeline networks. As the most vital part of risk-based inspections, risk assessment makes a significant contribution to achieving this aim. Accurate assessment can target high risk inspected regions so that limited resources can mitigate considerable risks in the face of increased spatial distribution of a pipeline network. However, the existing approaches for risk assessment face grave challenges due to a lack of sufficient data and an assessment’s vulnerability to human biases and errors. This paper attempts to tackle those challenges through spatial statistics, which is used to estimate the uncertainty of risk based on a dataset of locations of pipeline network failure events without having to acquire additional data. The consequence of risk in each inspected region is measured by the total cost caused by the failure events that have occurred in the region, which is also calculated in the assessment. Then, the risks of the different inspected regions are obtained by integrating the uncertainty and consequences. Finally, the feasibility of our approach is validated in a case study. Our results in the case study demonstrate that uncertainty is less instructive for prioritizing pipeline inspections than the consequences of risk due to the low significant difference in risk uncertainty in different regions. Our results also have implications for understanding the correlation between the spatial location and consequences of risk.

show abstract

Section: Resultsmentioning

confidence: 99%

A Spatial Statistic Based Risk Assessment Approach to Prioritize the Pipeline Inspection of the Pipeline Network

Hou

Dong

2020

Energies

View full text Add to dashboard Cite

show abstract

“…Formal models apply the probabilistic approach for assess the scenarios of risk with the use of Bayesian networks, being representative of the Why Because Analysis [86] method; fuzzy logic, Monte Carlo analysis, and Delphi procedure are additionally applied in the analysis of a container shipping logistic platform, and in a gas storage facility [146][147][148]. Industry 4.0 is a generic concept to improve self-control and risk identification through neural networks and machine learning; related to its application in industrial parks, it is used in inspection maintenance, construction, and environmental protection for chemical, oil and gas, and energy processes [149][150][151] Safety Barrier models. The representatives for this group are the Process Hazard Prevention Accident Models (PHPAM) and the System Hazard Identification Prediction and Prevention (SHIPP).…”

Section: Formal Basedmentioning

confidence: 99%

Analysis and Characterization of Risk Methodologies Applied to Industrial Parks

et al. 2020

View full text Add to dashboard Cite

It is important to evaluate the risks in industrial parks and their processes due to the consequences of major accidents and especially the domino effect. Scientific works present a wide possibility of models to deal with these situations. In this work, based on the information extracted from the scientific literature, six groups of risk methodologies are defined, analyzed, and characterized with methods that cover the standards, preventive, probabilistic, traditional, modern, and dynamic evaluation that are applied or could be used in industrial parks. It also tries to achieve the objective of determining which are more appropriate if the possible situations and causes that can produce an accident are taken into account, identifying and evaluating them with characteristics of simultaneity and immediacy, determining the probability of an accident occurring with sufficient advance in time to avoid it under the use of a working operational procedure. There is no definitive methodology, and it is necessary that they complement each other, but considering the proposed objective, the integrated application of traditional methodologies together with the management of safety barriers, the dynamic evaluation of risks, and the inclusion of machine learning systems could fulfill the proposed objective.

show abstract

“…As a non-parametric ensemble method, random forests (RFs) has gained popularity in dealing with regression and classification problems [ 28 , 29 ]. The RFs develops many decision trees [ 30 , 31 ] using a random subset of variables obtained independently and with replacement from the original dataset [ 27 , 32 ]. One of the important features of the RFs is the built-in feature importance functionality that helps rank the independent variable regarding their importance to the outcome variable, which adds value in data analysis [ 27 , 33 ].…”

Section: Introductionmentioning

confidence: 99%

Exploring drivers of patient satisfaction using a random forest algorithm

Simsekler¹,

Alhashmi²,

Azar³

et al. 2021

BMC Med Inform Decis Mak

View full text Add to dashboard Cite

Background Patient satisfaction is a multi-dimensional concept that provides insights into various quality aspects in healthcare. Although earlier studies identified a range of patient and provider-related determinants, their relative importance to patient satisfaction remains unclear. Methods We used a tree-based machine-learning algorithm, random forests, to estimate relationships between patient and provider-related determinants and satisfaction level in two of the main patient journey stages, registration and consultation, through survey data from 411 patients at a hospital in Abu Dhabi, UAE. Radar charts were also generated to determine which type of questions—demographics, time, behaviour, and procedure—influence patient satisfaction. Results Our results showed that the ‘age’ attribute, a patient-related determinant, is the leading driver of patient satisfaction in both stages. ‘Total time taken for registration’ and ‘attentiveness and knowledge of the doctor/physician while listening to your queries’ are the leading provider-related determinants in each model developed for registration and consultation stages, respectively. The radar charts revealed that ‘demographics’ are the most influential type in the registration stage, whereas ‘behaviour’ is the most influential in the consultation stage. Conclusions Generating valuable results, the random forest model provides significant insights on the relative importance of different determinants to overall patient satisfaction. Healthcare practitioners, managers and researchers can benefit from applying the model for prediction and feature importance analysis in their particular healthcare settings and areas of their concern.

show abstract

Machine learning approach for risk-based inspection screening assessment

Cited by 83 publications

References 82 publications

A Spatial Statistic Based Risk Assessment Approach to Prioritize the Pipeline Inspection of the Pipeline Network

A Spatial Statistic Based Risk Assessment Approach to Prioritize the Pipeline Inspection of the Pipeline Network

Analysis and Characterization of Risk Methodologies Applied to Industrial Parks

Exploring drivers of patient satisfaction using a random forest algorithm

Contact Info

Product

Resources

About