房地產估價對於一般民眾、銀行、政府及法院來說，都是非常重要的議題，對於這個問題如果要使用個別估價所需要消耗的人力、時間及資源會相當龐大，過往不動產估價模式相關研究，大多使用市場比較法、特徵價格法、決策樹及類神經網路等。本研究方法整合上述各方法的優點，先使用市場比較法來選取出與勘估標的條件相似之比較標的，再使用可以自動調整的特徵價格法或類神經網路進行模型建構；實驗顯示，本研究建構之模型能有效提升預測之精準度。

The real estate appraisal is an important issue for the general public, banks, government and courts. If such issue is resolved by using the individual real estate appraisal, it may be spent a lot of manpower, time and resources. In the past, many researches used sales comparison approach, hedonic price method, decision tree and neural network to predict the house prices. This study combines the advantages of the above methods. The sale comparison approach is used to select other subjects which are similar to target subjects for comparison. The hedonic price method and neural network model are used to construct a model for predicting the house prices. The experimental results show that the proposed model can improve the accuracy of predictions and have more effective than other methods proposed in the past researches.