Data-driven recognition and modelling of deterioration patterns in the US National Bridge Inventory: A genetic algorithm-artificial neural network framework

•Large-scale dataset assembled with information for 223,369 US bridges.•Essential bridge structural deterioration factors sought among 53 explanatory variables.•ANN pattern recognition framework introduced for capturing bridge deterioration information.•GA-ANN cascade optimization procedure optimally identifies ANN architecture and critical explanatory variables.•GA-ANN procedure minimizes the prediction error of bridge structural condition/deterioration status.