随着房地产市场的发展,房地产估价的研究逐渐深入。如何更好地、科学地、快速地对大量房地产价格进行评估已成为现实的问题。一般的平差模型对已有数据的利用率不高,当存在多类观测值时,不同观测值的随机模型的确定是否合理会严重影响结果的精度。基于房地产整体估价的思想,提出了方差分量估计建立价格体系的新方法。通过对两类观测值进行试验,一类通过比价系数求得价格差建立观测方程;一类由价格直接建立观测方程。结果分析表明,新方法能够实现不同观测值的权值自动调整,精度更高,可靠性更强。
With the development of China' s real estate market, the studies on real estate valuation are more deeply. It' s a real- istic problem to evaluate enormous real estate prices better, scientific and faster. Because of the low utilization of the existing data by general adjustment model, the accuracy of the resuhs are seriously affected by the weights of observation, when there are many types of observations. Based on the theory of the overall real estate valuation, we propose a new method establish price systems with variance component estimation. For comparison, two types of observations used in this paper in which, one establishes observation equation using the price difference obtained by price coefficient, while the other establishes observation equation directly by the price. The results show that the new method can auto revise the weights of the different observations and the results become more precise and reliable.