Development of EstiMatik: Artificial Neural Network-Based Tool for Predicting Construction Material Prices Based on Macroeconomic Indicators Using Python
摘要
In the National Capital Region (NCR) of the Philippines last 2022, construction materials prices peaked in 14 years, with the Construction Materials Wholesale Price Index (CMWPI) rising by 8.3%, a significant jump from 3.2% in 2021. Masonry materials, including Portland Cement (PC), and tinsmithry materials such as Reinforced Steel Bar (RSB), have surged annual increases of 123% and 141% since 2012, respectively. The cost increase of construction materials is crucial, as it accounts for 65%–75% of project costs, leading to cost overruns where actual costs exceed initial cost estimates by as much as 183%. Macroeconomic factors including Gross Domestic Product (GDP), Money Supply, Interest Rate, Consumer Price Index (CPI), Unemployment Rate, Producer Price Index (PPI), Inflation Rate, Exports, US Dollar Exchange Rate, and Lending Rate influence these price hikes, accurate prediction tools become essential for cost management. Using Python programming, the research aims to develop EstiMatik, a computational tool designed using Artificial Neural Networks (ANN) to forecast construction material prices about key macroeconomic indicators in the NCR. By compiling historical price trends and correlating them with macroeconomic indicators, EstiMatik enables contractors, suppliers, and government to anticipate future price changes and make informed decisions, therefore minimizing cost overruns. Stakeholders can improve budgeting, optimize resource allocation, and better manage financial risks in construction projects.