NEURO-FUZZY MODELS FOR CONSTRUCTABILITY ANALYSIS

Abstract

With the emergence of the new computer science areas of artificial intelligence and neural networks, researchers have applied them in the construction industry successfully. This paper presents comparative studies of two machine learning models namely backpropagation (BP) and Fuzzy ARTMAP based neuro-fuzzy models for handling qualitative fuzzy information of constructability evaluation. These models not only perform like traditional machine algorithms, but also handle missing information with better accuracy. Performance evaluation of the network has been carried out using traditional statistical tests. From the study, it was found that the Fuzzy ARTMAP model performs much better than the BP model.