Application of Projection Pursuit Model in residential building comprehensive performance assessment

To assess residential building comprehensive performance involves a number of indices, each index weight is difficult to identify and the result is tend to be interfered by subjective factors. Therefore, the Projection Pursuit Classification Model (PPC) and a Real coded Accelerating Genetic Algorithm (RAGA) are presented to convert the multi-dimensional problem to one-dimensional subspace and solve the multi-dimensional global optimization problem. According to the projection value to asses the comprehensive performance of residential building. It is easy to make decision, the result is relatively objective, and the projection values are dispersed. PPC provides a new approach to assess the comprehensive performance of residential building.