Performance Prediction Method in the Early Design Stage for Outdoor Ventilated Crawl Spaces Based on Artificial Neural Networks

Artikel i vetenskaplig tidskrift, 2010

The purpose of this article is to explore the possibility of using a tool based on artificial intelligence and real-life data. The aim is to develop and analyze one artificial intelligence method for one design part of a single-family house. Real-life data from documented experiences have been used as training data to develop a neural network to predict the performance of a specified design part, in this case, the outdoor ventilated crawl space. The results of this study indicate that this is an approach that could usefully be developed and investigated further. The tool managed to predict smell 100%, mold 76%, and rot 92% correctly.