Applying Decision Tree Algorithms to Develop Go/No Go Decision Model for Owners

Abstract

Go/No-Go execution decision is considered as the most important strategic decision for owners and project management consultants. This decision must be analyzed during the early stages of the project. Restructuring the process of decision-making may have positive results on the stability of the owner in the construction industry for longer term. The purpose of this study is to establish a proper go/no-go decision tree models for owners. The methodology of the models was developed using Exhaustive Chi-square Automatic Interaction Detector (Exhaustive CHAID) and Quick, Unbiased, Efficient Statistical Tree (QUEST) algorithms. Twenty-three go/no-go key factors were collected through extensive literature review. The go/no-go factors were listed based on their importance index as a result of a questionnaire contacted and distributed among the construction professionals. These factors were divided into four main risk categories; namely, organizational, project/technical, legal and financial/economic which are considered as inputs for models. Split-sample validation was applied for testing and measuring the accuracy of the Exhaustive CHAID and QUEST models. Moreover, Spearman’s rank correlation and Analysis of variance (ANOVA) tests were employed to identify the statistical features of the received 100 responses. Through extensive comprehensive literatures of previous studies conducted,this thesis contributes to the literature in three ways. First, it addresses the gap in literature by reviewing the current practice for conducting feasibility analysis used in construction project and studying the existing go/no-go models. Second, it provides list of decision supporting tools used in construction project and its limitation. Lastly, it is worthwhile to identify the most factors that affecting owner’s decision making. Accordingly, findings from this study set out a potential set of benchmarks for companies to use when deciding the criteria to be employed to evaluate new construction project. Another contribution is to propose a Go/No Go model which will support owner’s decision in the early stage before the project commences by applying decision tree algorithms QUEST and exhaustive CHAID. The model will evaluate anticipated risk factors in the project and reduce level of uncertainty in addition to simplifying decision making for owner away from complicated mathematical method.