基于粗集支持向量机的房地产预警系统分析-analysis of real estate early warning system based on rough set support vector machine.docxVIP

优秀毕业论文 精品参考文献资料 摘 要 房地产预警系统，是以有关科学理论和经济运行规律为指导，对房地产业运行历史过程的 经验分析。作为经济预警在房地产行业的应用，房地产预警系统主要通过指标体系的选择和建 立，揭示和认识具体指标与房地产业运行总体态势和局部特征的内在联系，在此基础上对潜在 地问题进行分析和采取对策，并对其走势做出正确的预测与评价，以期为政府及相关单位提供 决策支持，从而促进行业的持续健康发展。 本文在对各种经济预警方法比较分析的基础上，首先参考以往文献在影响房地产市场运行 的众多指标中进行初选，其次在若干初选指标中进行筛选，在筛选过程中首先采用定性与定量 相结合的分析方法选取两个在房地产行业最具代表性的指标，将其作为决策性指标，接着采用 粗糙集方法对初选的指标进行数据预处理和指标约简，最终确定影响房地产行业的重要指标并 建立房地产预警指标体系。另外本文对房地产预警警度划分成过冷、微冷、平衡、微热、过热 五个区间。 在建立基于粗糙集数据预处理的支持向量机预警模型的过程中，首先结合银川市的实际情 况，构建了银川市房地产预警系统指标体系。同时根据上述指标采集银川市不同时期的历史数 据，选择合适的数学方法、模型对预警指标进行测算，之后运用 3σ 警界控制原理对各状态的 区间进行确定，并以此划分出不同的警度，最后根据警限值将预测值转化为警度并进行预报。 本文运用基于粗集的属性归约方法对不同时期的历史数据进行预处理，并用支持向量机预测模 型对 2006-2007 年房地产行业的警情指标进行测算，最后根据警度对房地产市场景气状况进行 预测。 关键字：支持向量机，房地产，预警，粗糙集 宁夏大学硕士学位论文 Abstract Abstract Real estate early-warning system, which belongs to category of economical warning, is the empirical analysis of the real estate industry’s historical process, under the guidance of related scientific theory and economic operating laws. It is mainly reveals the internal relationships between the specific indicator with overall situation and the local feature of the real estate industry, and then accurately judgment of the above indicators could be obtained then. And it analyzes the potential problems and takes the countermeasures, makes a correct prediction and evaluation of the movements, so that it can take control measures and promote continuous and healthy operation of the real estate industry. In the process of establishing the early-warning index system, the primary indicators were chosen from the ones which affected the real estate market according to the reference literature. Then filter the primary indicators, there were two most representative indicators in the real estate industry had been selected as the decision-making indicators, with the analyzing method combined of qualitative and quantitative. After that, the data preprocessing and targets reduction for the primary indicators have been done by the method of Rough Set.. Ultimately,