城市住房价格PSO-LSSVR预测模型研究

发布时间：2018-04-03 23:55

本文选题：房地产市场　切入点：住房成交量　出处：《重庆大学》2014年博士论文

【摘要】：近年来，房地产行业的迅速蓬勃发展是大家有目共睹的。而城市商品住房作为其重要组成部分的也呈现了稳步增长的趋势。在住房产业快速发展的过程中，住房价格作为重要的经济杠杆对住房产业化与商品化起着重要的推动作用，住房价格也成为政府、居民和广大房地产开发商普遍关注的焦点。尤其是近年来，持续走高的城市住房价格给城市居民以及整个社会的经济发展都带来了很大的负面影响。与此同时，政府宏观调控措施虽然取得了一定的成效，但是效果不甚明显。基于此，本文将展开对城市住房价格预测模型的研究。通过建立价格预测模型，可以掌握房地产价格走势，合理准确评估预测房价，进而可以对房地产市场的发展展开分析，这对确保我国住房市场稳定健康发展有着重要的作用。本文的主要研究内容如下： ①本文首先梳理了国内外关于住房价格预测的相关研究，并基于已有的研究提出了基于粒子群算法的最小二乘支持向量机（PSO-LSSVR模型）的住房价格预测方法。最小二乘支持向量机在建模数据过程中能很好弥补人工神经网络模型和支持向量机的诸多不足，而且粒子群算法能迅速快捷地对参数进行优化，具有精度高、速度快等优点。 ②通过确定房地产度量指标体系和等级划分标准，详细介绍城市住房价格PSO-LSSVR预测模型的运作流程。并介绍基于PSO-LSSVR模型与模糊灰色理论的房地产市场预测系统架构，通过预测系统构架可以使预测模型更好的发挥作用。 ③以北京市为例展开预测模型的实证分析。通过构建相应的PSO-LSSVR住房价格和成交量预测模型，对北京市房地产市场发展健康状况进行评估。实证分析表明基于粒子群优化最小二乘支持向量回归预测模型优于传统的支持向量回归模型，也证明了PSO-LSSVR预测模型用于城市住房价格预测的有效性。 ④在住房价格、住房成交量的PSO-LSSVR预测模型和基于模糊灰色理论的房地产市场评估模型基础上给出了整个房地产市场预测系统的实现过程，并详细介绍系统的软件实现过程。 ⑤总结本研究的研究成果并对未来的研究提出建议。本文提出城市住房价格PSO-LSSVR预测模型，并结合模糊灰色理论提出一整套房地产市场预测与评估方法，对于房地产市场预测与评估有着非常重要的价值。通过该研究既可以为各国家及城市政府管理部门制定房地产调控政策，保证房地产经济健康、持续、稳定的发展提供重要的手段和决策依据，，可以为企业的投资决策提供更多的帮助；也可以更进一步完善我国房地产研究的理论系统，具有重要的理论意义和实践意义。
[Abstract]:In recent years, the rapid and vigorous development of the real estate industry is obvious to all.As an important part of urban commercial housing, it also shows a steady growth trend.In the process of rapid development of housing industry, housing price, as an important economic lever, plays an important role in promoting housing industrialization and commercialization. Housing price has become the focus of the government, residents and real estate developers.Especially in recent years, rising urban housing prices have brought great negative effects to urban residents and the economic development of the whole society.At the same time, although the government macro-control measures have achieved some results, but the effect is not obvious.Based on this, this paper will carry out a study on the urban housing price prediction model.Through the establishment of price forecasting model, we can grasp the trend of real estate price, reasonably and accurately evaluate and forecast the housing price, and then analyze the development of real estate market, which plays an important role in ensuring the stable and healthy development of our country's housing market.The main contents of this paper are as follows:Firstly, this paper reviews the research on housing price prediction at home and abroad, and proposes a Particle Swarm Optimization (PSO) based LS-LSSVR model for housing price prediction.The least square support vector machine (LS-SVM) can make up for many shortcomings of artificial neural network model and support vector machine in the process of modeling data, and particle swarm optimization algorithm can quickly and quickly optimize the parameters, which has the advantages of high precision and fast speed.(2) by determining the real estate measurement index system and grade classification standard, the operation process of urban housing price PSO-LSSVR forecasting model is introduced in detail.The structure of real estate market forecasting system based on PSO-LSSVR model and fuzzy grey theory is introduced.3 taking Beijing as an example, the empirical analysis of forecasting model is carried out.By constructing the corresponding PSO-LSSVR housing price and volume forecasting model, this paper evaluates the health of the real estate market in Beijing.The empirical analysis shows that the least square support vector regression forecasting model based on particle swarm optimization is superior to the traditional support vector regression model, and it also proves the validity of PSO-LSSVR forecasting model in urban housing price forecasting.4. Based on the PSO-LSSVR forecasting model of housing price, housing transaction volume and the real estate market evaluation model based on fuzzy grey theory, the realization process of the whole real estate market forecasting system is given, and the software realization process of the system is introduced in detail.5 summarize the research results of this study and put forward suggestions for future research.This paper puts forward the PSO-LSSVR forecasting model of urban housing price, and puts forward a set of real estate market forecasting and evaluation methods combined with fuzzy grey theory, which has very important value for real estate market prediction and evaluation.Through the research, it can provide important means and decision basis for the governments of various countries and cities to formulate the real estate regulation and control policies and ensure the healthy, sustainable and stable development of the real estate economy.It can provide more help for the investment decision of enterprises and perfect the theoretical system of real estate research in our country, which has important theoretical and practical significance.
【学位授予单位】：重庆大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：F299.23

【参考文献】