Development of a decision support system for LEED for EB credit selection based on climate factors

Abstract

© ASCE 2014. LEED is a credit-based rating system that provides a third-party verification of green buildings worldwide. A building can obtain a Platinum, Gold, Silver or Certified grade based on the number of LEED credit points achieved. Selection of target credits is important and challenging for LEED managers due to limited budget, tight project schedule, and limited resources in many green building projects. Local climate factors like temperature can affect the selection of green building technologies and hence the LEED credits adopted. However, no research has been done to suggest LEED target credits based on climate factors. This paper aims to develop a decision support system based on climate factors for LEED credit selection using data mining techniques. The LEED for Existing Buildings version 2009 was focused in this study. Information of 912 certified green building projects and their surrounding climate circumstances was collected and studied. Classification models for 48 LEED credits that use credit achievement as the class and climate factors as the variables were then constructed and optimized using three data mining algorithms - Random Forests, AdaBoost Stumps and Support Vector Machine (SVM). The results were incorporated in a web-based decision support system. A case study was then conducted to illustrate and evaluate the system. The results showed that our decision support system has a high accuracy.