English  |  正體中文  |  简体中文  |  Items with full text/Total items : 43312/67235
Visitors : 2107302      Online Users : 17
RC Version 5.0 © Powered By DSPACE, MIT. Enhanced by NTU/NCHU Library IR team.

Please use this identifier to cite or link to this item: http://nchuir.lib.nchu.edu.tw/handle/309270000/151616

標題: 類神經網路應用於法拍不動產估價
Artificial Neural Network on Court Auction Houses
作者: 劉時旭
Liu, Shin-Hsu
Contributors: 謝孟勳
Machine Hsie
土木工程學系所
關鍵字: 類神經網路;法拍不動產;法拍屋;不動產估價;超商數
Artificial Neural Network;Court Auction Real Estate;Court Auction Houses;Real Estate Appraisal;Number of Chain Convenience Store
日期: 2012
Issue Date: 2013-11-07 11:20:22 (UTC+8)
Publisher: 土木工程學系所
摘要: 法拍屋乃是法院進行拍賣因債務抵押之不動產。一般而言,拍定價格會比市場行情價格為低,因而在房地產價格越來越高的近年來,吸引了大量投標者進入法拍屋市場。例如:單單就我國2010年法拍市場統計,法拍件數6萬630件,拍定金額高達1344億元。

法拍市場在近年來逐漸受到人們重視,然而目前針對法拍不動產估價做為研究之成果仍極度缺乏,由於不動產市場為一不完全市場,所得之資訊常會有不完整之處,對不動產的價格也僅能有一些基本概念性的推估,如果能從這些資訊中推估出較為準確的價格,將有助於降低買方的風險。因此本研究,應用類神經網路建立一個法拍不動產拍定價格預估模式,期能提供買方使用者做一參考使用。本研究收集法拍屋拍定價格之實際案例,並篩選評估可取得性,精挑重要資訊作為輸入變數,包含:(1)面臨路寬、(2)基地面臨道路數、(3)人口密度(流量)、(4)是否點交、(5)產權持份類型、(6)公告現值、(7)土地持份、(8)樓地板面積、(9)拍定價格、(10)第幾次拍賣,等數個變數。分析檢討與各項相關的輸入變數,透過類神經網路的學習訓練及參數改進修正,最後進行法拍不動產拍定價格預估。其中,「人口密度」此一選項,一般民眾無法有效取得此一資訊。本研究權宜採用:方圓五百公尺內之超商數目做為一分析指標。此一創新且便利準確之輸入變數,大幅提升研究成果之精確性。研究成果顯示:類神經網路確實可得到快速、精確的預估成果,建議非常適合作為法拍屋購買的決策評估使用。
Auction houses are the real estate of the Foreclosure. Generally speaking, the auction price is lower than general market price. In recent years, the real estate price getting higher and higher, and it attracts a large number of bidders entering the foreclosure market. For example, in 2010, the market statistics number for foreclosure is 60630, the amount of money is NT.134.4 billion .

In recent years, auction houses had been subjected to people''s attention. But the information for auction houses is still extremely lacking. Due to the real estate market is an imperfectly competitive market, the proceeds of the information is often not be complete. The price can only be estimated by some fundamental concepts. By using the information to estimate a more accurate price will help reducing the risk for buyers. Therefore, this study use Artificial Neural Network to establish a model to estimate auction houses .By collecting cases of auction houses, we select the finest information as input variables, including: (1)the width of facing road (2)numbers of facing road (3) population density, (4)delivered by the court or not, (5) property rights, (6) current assessed land value, (7) land possession, (8) floor area, (9) bid price, (10) how many times the auction, etc. Analyze and review the relevant input variables Improvement Amendments through the training and the parameters of the artificial neural network learning, we can estimate the price. In particular, the population density is often that can’t be effectively achieved. In this study, we use the number of chain convenience store within a radius of 500 meters as an analysis of indicators. This innovative and facilitate accurate input variables, significantly increasing the accuracy of the research results. The research results show that: the artificial neural network is indeed available fast, accurate forecast of the result, so we recommend that this is good for assessment of auction houses.
Appears in Collections:[依資料類型分類] 碩博士論文

Files in This Item:

File Description SizeFormat
nchu-101-5099062020-1.pdf1330Kb335View/Open
index.html0KbHTML134View/Open


 


學術資源

著作權聲明

本網站為收錄中興大學學術著作及學術產出,已積極向著作權人取得全文授權,並盡力防止侵害著作權人之權益。如仍發現本網站之數位內容有侵害著作權人權益情事者,請權利人通知本網站維護人員,將盡速為您處理。

本網站之數位內容為國立中興大學所收錄之機構典藏,無償提供學術研究與公眾教育等公益性使用。

聯絡網站維護人員:wyhuang@nchu.edu.tw,04-22840290 # 412。

DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU/NCHU Library IR team Copyright ©   - Feedback