Modified Stepwise Regression Approach to Streamlining Predictive Analytics for Construction Engineering Applications

Mohsenijam, Arash; Siu, Ming-Fung Francis; Lu, Ming

doi:10.1061/(asce)cp.1943-5487.0000636

Cited by 12 publications

(3 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Note that detailed results for MLR and MSR implementation and comparison on the same dataset can be found in Mohsenijam et al (2017); herein, the developed predictive models along with respective evaluation metrics are presented in Table 1 and 2. The "M5+MSR" modeling process in accordance with the proposed methodology is elaborated as follows:…”

Section: Model Developmentmentioning

confidence: 99%

See 1 more Smart Citation

Integrating model tree and modified stepwise regression in concrete slump prediction and steel fabrication estimating

Mohsenijam

Naumets

2022

Can. J. Civ. Eng.

Self Cite

View full text Add to dashboard Cite

The model tree algorithm of M5 is integrated with the multiple linear regression technique called modified stepwise regression (MSR), resulting in a new methodology for modeling complex civil engineering problems. We purposefully chose artificial neural networks (ANN) for comparison against the proposed “M5+MSR” because they fall at the two ends of the model interpretability spectrum in machine learning. In two application cases (case 1: concrete workability and case 2: steel fabrication estimating), the proposed “M5+MSR” gave rise to explainable regression tree models featuring substantially reduced complexities against ANN and model prediction errors comparable to ANN. In both cases, the resulting “M5+MSR” models consistently outperformed ANN in terms of model overfitting metrics by 19% in case 1 and by 21% in case 2, thus boasting better learning performances. The proposed new methodology will potentially find applications in tackling a wide range of complicated engineering problems that entail fitting prediction models based on laboratory or field data.

show abstract

Section: Model Developmentmentioning

confidence: 99%

“…Variable selection: the tree structure and the samples that reach the leaves resulting from the previous steps need to undergo MSR variable selection process [detailed steps can be found in (Mohsenijam et al 2017)].…”

Section: R a F Tmentioning

confidence: 99%

Integrating model tree and modified stepwise regression in concrete slump prediction and steel fabrication estimating

Mohsenijam

Naumets

2022

Can. J. Civ. Eng.

Self Cite

View full text Add to dashboard Cite

show abstract

“…It requires a comprehensive review during the development phase before the physical implementation of construction. Cost estimation significantly impacts the success of both the project and the company [8].…”

Section: Introductionmentioning

confidence: 99%

Project Cost Estimation Using a Stepwise Approach: a Case Study of an Infrastructure Project in Gresik Regency, East Java

Siswoyo,

Soepriyono,

Suharso

et al. 2023

JTSP

View full text Add to dashboard Cite

There are many studies related to cost estimation applied in a project, but only a few employ using “stepwise” method. This approach is expected to assist in solving complex problems by establishing a criteria hierarchy, subjectively assessed by interested parties, and incorporating various considerations to develop weights or priorities. The foundation of the stepwise approach relies on well-documented findings, which assert that 80% of the total project cost corresponds to the 20% most expensive components of the work. The research systematically collected data for project development estimates based on the bill of quantities from each year, as well as the HSPK (activity main unit price) in Gresik Regency from 2016 to 2018. This research aims to develop a conceptualized estimation model that provides clear, rapid, and relatively accurate initial project cost information. The results of this research include analysis of stepwise approach on cost realization, analysis of accuracy of the cost estimation model for buildings using the stepwise method, and an analysis of the cost of road and bridge infrastructure projects to facilitate completion in Gresik Regency. The research concludes that the analysis of the Cost Significant Model's impact on cost realization in the Public Works and Spatial Office's Highways Division in Gresik Regency has resulted in a cost estimation model for road development with an accuracy range from -12.7% to 16.0%, with an average accuracy rate of 0.59%. Additionally, there is a very strong correlation (R) of the independent variable (work in Division 2) and the total development cost, indicating that fluctuations in the cost of significant components are closely linked to changes in the overall development cost. The analysis of infrastructure project costs related to roads and bridges for project completion revealed that the highest value among independent variables is associated with Division 7, at Rp. 4,403,054 per m2. This suggests that the cost of Division 7 work is the most influential, while Division 1 (X1) has the lowest cost, at Rp. 35,589 per m2. Furthermore, there is a work component with zero cost, namely, Division 4 works.

show abstract

Application of Predictive Analytics in Built Environment Research: A Comprehensive Bibliometric Study to Explore Knowledge Domains and Future Research Agenda

Halder

Batra

2023

Arch Computat Methods Eng

View full text Add to dashboard Cite

Modified Stepwise Regression Approach to Streamlining Predictive Analytics for Construction Engineering Applications

Cited by 12 publications

References 27 publications

Integrating model tree and modified stepwise regression in concrete slump prediction and steel fabrication estimating

Integrating model tree and modified stepwise regression in concrete slump prediction and steel fabrication estimating

Project Cost Estimation Using a Stepwise Approach: a Case Study of an Infrastructure Project in Gresik Regency, East Java

Application of Predictive Analytics in Built Environment Research: A Comprehensive Bibliometric Study to Explore Knowledge Domains and Future Research Agenda

Contact Info

Product

Resources

About