Predicting the impact of climate change and socio-economic parameters on infrastructure projects using artificial intelligence: The case of Adana province

The scale and complexity of infrastructure projects are critically important for meeting the long-term needs of society. These projects aim to establish sustainable and reliable infrastructure systems, ensuring a livable environment for future generations. In this thesis, machine learning algorithms were employed to examine the impacts of climate change and socio-economic parameters on drinking water and wastewater infrastructure projects in Adana Province. To determine the parameters to be used in the models, SWARA (Step-wise Weight Assessment Ratio Analysis) and PCA (Principal Component Analysis) methods were applied. Based on data from 2006 to 2023, Lasso, Ridge, XGBoost, and SVR were used to forecast distribution flow rates and, consequently, the future scale of drinking water and wastewater projects. The performance of the developed models was evaluated using MAE, RMSE, R² and Nash-Sutcliffe Efficiency (NSE) metrics. The results demonstrated that three socio-economic parameters urban development, population, and registered subscribers along with the Aydeniz Drought Index (a combination of temperature, sunshine duration, precipitation and humidity), significantly influenced the cost components. These findings emphasize the importance of considering such factors in the planning and design processes of infrastructure projects.