Model of Predicting the Rating of Bridge Conditions in Indonesia with Regression and K-Fold Cross Validation

Maintenance and repair of the bridge are inevitable in operation a to maintain its condition keep operation. Indonesia has hundreds thousands bridges that still actively use. The classic problem with infrastructure management, such as bridges, is large numbers generally not balanced adequate maintenance budgets. Therefore,the strategy implementing by preparing priorities becomes only logical approach. To get priority scale, scoring mechanism needed. assessment used Ministry Public Works Housing (PUPR) especiallythe Bina Marga field based on management system, namely Bridge Management System (BMS) 1993. With BMS 1993, represented Condition Value (NK) bridge. This study existingNK, prediction NK value future. predicted model developed regression models. Regression models combined k-fold cross-validation improve accuracy rate model. produces for all variables values low error percentage range MAPE = 10% RMSE 0.15. Further significance tests ANOVA also conducted test effect independent dependent variables, including testing fit show resulting does overfit and/or underfitting.