本文選題：房地產(chǎn)市場預(yù)警　切入點：人工神經(jīng)網(wǎng)絡(luò)　出處：《東北財經(jīng)大學(xué)》2013年碩士論文　論文類型：學(xué)位論文

【摘要】：隨著住房制度改革和城市化進程的加快,我國房地產(chǎn)業(yè)迅速發(fā)展,對我國國民經(jīng)濟的影響也越來越大。房地產(chǎn)業(yè)的蓬勃發(fā)展對改善居民生活、拉動經(jīng)濟增長發(fā)揮了重要作用。然而,房地產(chǎn)投資額的大幅增長也使得房地產(chǎn)業(yè)出現(xiàn)了一系列問題。近年來,我國房價逐年上漲,市場風(fēng)險逐步加大,關(guān)于房地產(chǎn)泡沫的討論越來越激烈。因此,建立城市房地產(chǎn)預(yù)警系統(tǒng),對我國房地產(chǎn)市場的發(fā)展狀況進行準確的判斷和監(jiān)控,對確保房地產(chǎn)業(yè)持續(xù)健康發(fā)展意義重大。 房地產(chǎn)預(yù)警是經(jīng)濟預(yù)警的組成部分,本文首先總結(jié)了國內(nèi)外關(guān)于經(jīng)濟預(yù)警的研究動態(tài),然后分析了我國房地產(chǎn)預(yù)警領(lǐng)域的研究現(xiàn)狀及存在的主要問題。通過對現(xiàn)有預(yù)警方法的優(yōu)缺點進行對比分析,鑒于人工神經(jīng)網(wǎng)絡(luò)具有非線性、容錯性、操作簡便等優(yōu)點,本文決定采用人工神經(jīng)網(wǎng)絡(luò)預(yù)警方法構(gòu)建本文的房地產(chǎn)預(yù)警模型。在大量閱讀整理相關(guān)文獻的基礎(chǔ)上,從房地產(chǎn)業(yè)發(fā)展速度、房地產(chǎn)業(yè)與國民經(jīng)濟的協(xié)調(diào)關(guān)系、房地產(chǎn)業(yè)內(nèi)部協(xié)調(diào)關(guān)系三個方面選擇了最常用、代表性較強的指標構(gòu)建本文的房地產(chǎn)預(yù)警指標體系�；谌斯ど窠�(jīng)網(wǎng)絡(luò)的房地產(chǎn)預(yù)警模型的構(gòu)建思路如下：首先,根據(jù)研究對象特點選擇警情指標,判斷以往年份的房地產(chǎn)市場警情,本文采用的是歸一化方法。然后,確定警兆指標,警兆指標應(yīng)當對警情指標的領(lǐng)先性,還要考慮到數(shù)據(jù)的可獲得性等因素。最后,采用Matlab軟件編寫了人工神經(jīng)網(wǎng)絡(luò)的訓(xùn)練程序?qū)颖緮?shù)據(jù)進行訓(xùn)練,得出我國房地產(chǎn)市場的預(yù)警模型。該預(yù)警模型經(jīng)過檢驗后即可運用于預(yù)警實踐。 為了對我國房地產(chǎn)市場整體運行狀況有一個更為全面的了解,本文按照經(jīng)濟發(fā)展程度的差別分別選取了北京、重慶、威海作為我國一二三線城市的代表城市進行實證研究,通過數(shù)據(jù)收集,分別建立了人工神經(jīng)網(wǎng)絡(luò)的訓(xùn)練樣本,并對預(yù)警結(jié)果的準確性進行了檢驗,然后將2012年房地產(chǎn)市場數(shù)據(jù)處理后得出的警兆指標輸入預(yù)警模型對下一年的房地產(chǎn)警情進行判斷,預(yù)警結(jié)果顯示北京市和威海市2013年房地產(chǎn)市場將處于“熱”的警情狀態(tài)下,2013年重慶市房地產(chǎn)市場將處于“正常”的警情狀態(tài)下。 在得出預(yù)警結(jié)論的基礎(chǔ)上,文章最后簡單分析了出現(xiàn)這種結(jié)果的原因,政府可從規(guī)范土地管理制度,調(diào)節(jié)土地供應(yīng)、控制房地產(chǎn)信貸規(guī)模、調(diào)整住房供應(yīng)結(jié)構(gòu)、限制房地產(chǎn)市場投機和炒作行為、增加房地產(chǎn)市場信息透明度等方面加強對房地產(chǎn)市場的管理。 由于我國房地產(chǎn)預(yù)警研究還不夠成熟,多數(shù)預(yù)警方法還僅僅停留在理論層面,距離實際應(yīng)用還有較多問題,盡管人工神經(jīng)網(wǎng)絡(luò)預(yù)警方法有其自身優(yōu)越性,是較為先進的科學(xué)方法,但我國房地產(chǎn)數(shù)據(jù)不足、關(guān)于房地產(chǎn)預(yù)警指標體系研究不夠成熟等因素也會使得該方法的運用受到較大局限。在后續(xù)的研究中還需要在這些方面有所加強。
[Abstract]:With the reform of housing system and the acceleration of urbanization, the real estate industry in our country is developing rapidly, and the influence on our national economy is becoming more and more great. Promoting economic growth has played an important role. However, the substantial increase in real estate investment has also caused a series of problems in the real estate industry. In recent years, housing prices in China have increased year by year, and the market risks have gradually increased. The discussion on the real estate bubble is becoming more and more intense. Therefore, it is of great significance to establish the urban real estate early warning system to accurately judge and monitor the development of the real estate market in our country, which is of great significance to ensure the sustained and healthy development of the real estate industry. Real estate early warning is an integral part of economic early warning. This paper first summarizes the research trends of economic early warning both at home and abroad. Then it analyzes the current research situation and main problems in the field of real estate early warning in China. By comparing and analyzing the advantages and disadvantages of the existing early warning methods, in view of the advantages of artificial neural network, such as nonlinear, fault-tolerant, easy to operate, etc. This article decides to use the artificial neural network early warning method to construct the real estate early warning model of this paper. On the basis of a large number of reading and sorting relevant literature, from the real estate industry development speed, real estate industry and national economy coordination relations, Three aspects of the coordination relationship within the real estate industry choose the most commonly used, more representative indicators to build the real estate warning index system. The real estate warning model based on artificial neural network is constructed as follows: first, According to the characteristics of the object of study, select the alarm index, judge the real estate market alarm situation in the past years, this paper adopts a normalized method. Then, determine the warning indicators, warning indicators should be the lead to the warning indicators, Finally, the training program of artificial neural network is compiled by Matlab software to train the sample data. The early warning model of China's real estate market is obtained, which can be used in early warning practice after being tested. In order to have a more comprehensive understanding of the overall operation of China's real estate market, this paper selects Beijing, Chongqing and Weihai as the representative cities of the 123 tier cities in China according to the difference of economic development. Through data collection, the training samples of artificial neural network are established, and the accuracy of early warning results is tested. Then input the warning indicators of the 2012 real estate market data processing into the early warning model to judge the real estate police situation in the next year. The warning results show that the real estate market in Beijing and Weihai will be in a "hot" state on 2013, and that in Chongqing on 2013, the real estate market will be in a "normal" state of warning. On the basis of the conclusion of early warning, the paper analyzes the reason of the result. The government can regulate the land supply, control the scale of real estate credit, adjust the structure of housing supply. To restrict speculation and speculation in the real estate market and to increase the transparency of information in the real estate market, and so on, to strengthen the management of the real estate market. Because the research of real estate early warning in our country is not mature enough, most of the early warning methods are still only in the theoretical level, and there are still many problems from the practical application, although the artificial neural network early warning method has its own superiority. Is a more advanced scientific method, but the real estate data in our country are insufficient. The application of this method will also be limited by some factors, such as the immaturity of the research on the early warning index system of real estate, and it is necessary to strengthen these aspects in the follow-up research.
【學(xué)位授予單位】：東北財經(jīng)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2013
【分類號】：TP183;F299.23

【參考文獻】

相關(guān)期刊論文 前10條

1 劉為敏;;論房地產(chǎn)調(diào)控政策在中小城市失靈的原因及對策[J];企業(yè)導(dǎo)報;2009年04期

2 胡志濤;黃躍平;;我國商業(yè)銀行房地產(chǎn)貸款風(fēng)險分析及其防范[J];商場現(xiàn)代化;2009年02期

3 郭磊,王鋒,劉長濱;深圳市房地產(chǎn)預(yù)警系統(tǒng)研究[J];數(shù)量經(jīng)濟技術(shù)經(jīng)濟研究;2003年07期

4 趙黎明,賈永飛,錢偉榮;房地產(chǎn)預(yù)警系統(tǒng)研究[J];天津大學(xué)學(xué)報(社會科學(xué)版);1999年04期

5 師應(yīng)來;王平;;房地產(chǎn)預(yù)警指標體系及綜合預(yù)警方法研究[J];統(tǒng)計研究;2011年11期

6 陳心儀;徐佳;;房地產(chǎn)泡沫的成因分析[J];現(xiàn)代商業(yè);2009年09期

7 丁烈云,李斌;房地產(chǎn)市場預(yù)警調(diào)控系統(tǒng)的構(gòu)筑技術(shù)要點及流程設(shè)計[J];系統(tǒng)工程理論與實踐;2002年04期

8 韓立達;我國房地產(chǎn)預(yù)警系統(tǒng)的基本特征及方法[J];燕山大學(xué)學(xué)報(哲學(xué)社會科學(xué)版);2004年03期

9 李濤,伍建平;房地產(chǎn)泡沫的成因及防范措施探討[J];山東理工大學(xué)學(xué)報(社會科學(xué)版);2004年02期

10 陳夢璇;;我國房地產(chǎn)泡沫與銀行信貸擴張關(guān)聯(lián)分析[J];中國商界(下半月);2010年03期

相關(guān)碩士學(xué)位論文 前2條

1 馬迪;大連市房地產(chǎn)預(yù)警研究[D];大連理工大學(xué);2006年

2 鄭佳;房地產(chǎn)預(yù)警指標體系構(gòu)建與實證研究[D];吉林大學(xué);2008年

，

本文編號：1615988