Predict The Risk Factor of The Possibility of Death For Not Having A Bridge In Nepal By Using Multiple Regression Analysis

Abstract

Communication and transportation are important aspects of human life, but in Nepal, the geographically challenging topography and disconnected communities by more than 6000 rivers and rivulets, present great challenges in their daily lives for accessing basic health services, education, and household chores. Bridges are one of the major means of connecting the rural population in Nepal, but the lack of appropriate and safe means of crossing torrential rivers has resulted in increased suffering for local communities and developmental challenges for the actors involved in this sub-sector. The focus of this research is to technically analyse the potential bridge sites based on the major prioritization indices and to determine risk factors related to particular locations leading to solutions for the permanent transportation problem. The government has been trying its’ best in collaboration with various development actors to address the problem and help to reduce the risks related to potential loss of lives while crossing the mountainous terrain to various destinations. Many practices proposed or implemented have been proven to be risky, especially for children, men, women, and the elderly population. This research aims to establish a proven module to construct a trail bridge as a safer means of transport across the hills that would accelerate access to basic services such as education, healthcare, and provide people with new development prospects. The innovative idea involves supporting the construction of a trail bridge to support services such as education, medical facilities, and household chores. The data has been analysed using Multiple Regression Analysis (MRA). In this research, the authors predicted the risk factor of the possibility of death due to the lack of a bridge, which depends on the total population, total households, river type, width of the bridge, etc. The model was evaluated using 50-50%, 66-34%, and 80-20% train-test splits and 10-fold cross-validation and an accuracy of approximately 70% was achieved. The authors collected data by mobilizing local NGOs and informing the public through local radio to conduct a comprehensive study of nationwide bridge demand. The secondary source of data for post-bridge building assessment is extracted through the project management information system.