Road infrastructural development is the bedrock for the social and economic development of any nation. However many developing nations including Nigeria still face the challenge of bridging the gaps in road infrastructure due to the impact of risk on project objectives leading to cost and time overruns or outright abandonment of the projects. This study is focused on the assessment of risk impact on road projects using a neural network approach. We developed neural-risk assessment models (NRAMs) using Keras application program interface(API) with the aid of Python programming language for the assessment of significant risk impact on the cost and time performance of road projects. The prediction of risk impact on the cost performance using NRAM revealed a mean operational error (MOE) of 64.26% for the ELU-linear model (ELM) indicating low accuracy while in time performance a MOE of 10.18% was observed for the softsign-linear model (SSLM) which indicates high accuracy. NRAM will assist the project managers in evaluating and predicting the impact of risk on the cost and time performance of road projects. Also, NRAM could be used as a decision support tool in the delivery process of road infrastructural projects in Nigeria and other developing countries.
Introduction
Neural Network (NN)
NN for risk management
Activation function (AF)
Methodology
Data collection
Risk management
Development of NRAM
Model evaluation
Application of the model
Discussion of Results
Conclusion
Data Availability Statement
Acknowledgment
References