Forecasting liquidated damages using random forest regression model for highway construction projects in Region 02 [manuscript] / Kenneth L. Bacud, Ma. Angelica B. De Guzman, Lei Vanessa E. Tan.
by Bacud, Kenneth L., author.
Year: 2025| Item type | Location | Collection | Call number | Status | Date due | Barcode |
|---|---|---|---|---|---|---|
Academic Research
|
Academic Research Section | Academic Research | Civil 0167 2025 c.1 (Browse shelf) | Available | CIVIL0167 |
Thesis (B.S.) -- Cagayan State University, 2025.
Highway construction projects in Region II, Philippines, face frequent delays, leading to significant liquidated damages (LDs) that impose financial penalties and hinder project success. This study utilized a Random Forest Regression (RFR) model to predict LDs across 199 highway projects, using data from 2015-2018 and 2021-2024 from the Department of Public Works and Highways - Region II and Cagayan Third District Engineering Office, RFR outperformed Linear Regression and Decision Tree models, achieving the lowest Mean Squared Error (276.83 billion), Root Mean Squared Error (526,148.09), and the highest R-squared (0.7292), explaining 72.92% of LD variance. Visual analyses confirm RFR's precision, with residuals mostly within ±1e6 PHP. precision, Key delay drivers included Original_Duration_Days (0.6451) and Contract_Cost_PhP (0.2117). The RFR model offers reliable LD forecasting, enabling better budgeting, reduced disputes, and timely project delivery, supporting Region II's economic growth, which contributes 7% to the Philippines' GDP in 2023.
Keywords: liquidated damages, random forest regression, highway construction projects, feature importance, project delays, risk management
There are no comments for this item.