An Artificial Neural Network Model for Predicting Condominiums Operation and Maintenance Costs in Early Design Phase

Building operation and maintenance (O&M) costs are affected by design decisions made in early design phase. To assist developers and architects of condominium properties in effectively assessing the impacts of their design decisions made in early design phase on future O&M costs, an artificial neural network (ANN) O&M cost prediction model is developed. A regression cost prediction model was further developed as a benchmark model for testing the predictive accuracy of the ANN model. Six design attributes were identified as critical input factors to both models. 65 condominium properties were randomly selected as samples and their design and cost data collected. 55 of them were treated as training samples whose data were used to develop the ANN and regression models, and the remaining ten as test samples whose data used to compare and verify the performance of both models. The study results revealed that the ANN model delivers more accurate cost prediction results, with lower average absolute error around 7.2% and maximum absolute error around 16.7%, as compared with the regression model.