-
首页
-
教学项目展开 / 收起
sidenav background教学项目
sidenav header background[12月19日]发展经济学workshop
发布日期：2018-12-17 09:32 来源：北京大学国家发展研究院
时间：2018年12月19日（周三）14：00-15：30
地点：北大国发院/中国经济研究中心万众楼小教室
主持人：张晓波、李力行、席天扬
主讲人：张晓奇题目：Constrained clustering and two applications: detecting the impact of urban structure on the relationship between subway system and housing price & modeling the intra-urban housing market spillover
摘要：We develop a constrained clustering technique that has quite promising applications to topics in urban economics and housing market studies. I am going to present two applications of this method with different types of constraint.
The first application refers to detect the influence of urban structure on the relationship between subway system and housing prices, the following is the abstract:
Using housing market data of Beijing and Hangzhou, China, we conduct a case study to detect how the difference of urban structure can affect the relationship between subway system and housing prices. To quantify the characteristics of urban structure, we propose a constrained clustering method, which cannot only reveal the spatial heterogeneity of housing market, but also provides a link between the heterogeneity and the underlying urban structure. Applying constrained clustering to Beijing and Hangzhou, we find that the relationship between accessibility to metro stations and housing prices is weak and vulnerable, while the improvement of commuting efficiency, measured by a key variable – metro index, does have a robust connection to the metro premium on housing units. In particular, only a large metro index can associate with positive metro premium. Structural features, like the size of city core and the existence of multi-centres, influence the metro premium through affecting the value and spatial distribution of metro index. The evidence from Beijing and Hangzhou supports that in a mono-centric city, the size of city core is positively associated with metro index and metro premium, while in a polycentric city with a small city core, metro index tends to be lower in the core region and higher in the satellite regions, which enforces the metro premium to be negative in core while positive out core.The second application refers to modelling the intra-urban housing market spillover, the following is the abstract:
This study analyses the spillovers among intra-urban housing sub-markets in Beijing, China. Intra-urban spillover is at least as important as the widely-studied inter-urban spillover, because it is crucial to the inequality issue within a city which is always a main concern of city managers. Intra-urban spillover imposes methodological challenge. Unlike the inter-urban situation, the range of every submarket is not naturally defined, therefore it is not possible to directly evaluate intra-urban spillover by standard time-series models. A novel clustering technique, namely constrained clustering, is proposed to solve that issue. Using constrained clustering, spillover effect is formulated as a spatial Markov chain model, identification of submarkets as hidden states of Markov chain and estimation of the Markov transition matrix can be done simultaneously. Using a sampled day-by-day transaction dataset of second-hand apartments in Beijing from 2011 to 2017, constrained clustering detects that there exist quite robustly 16 submarkets within the urban-suburban area of Beijing, spillover effect transits among the 16 regions with the highest transition probability appearing in the overlapped region of urban core and Tongzhou district. This observation agrees with the new urban planning proposal announced in early 2012 and being continuously effective during the entire data collection period. In addition to identifying the impact of policy, we also analyses a variety of spillover “types” through regression analysis of Markov transition matrix and trajectory simulation of the spatial Markov chain. Our “type” analysis shows that the “ripple” form of spillover is no longer dominant at the intra-urban level, although it is believed to be the dominant type at inter-urban level in literature. Other types of spillover, such as the spillover due to the existence of price depressed regions, which are not significant at inter-urban level, but do play central roles at intra-urban level. This observation reflects the complexity of spillover dynamics in the intra-urban situation and the distinct driving-forces behind intra- and inter-urban spillovers.主讲人简介：张晓奇，浙江财经大学金融学院，讲师。2017年6月毕业于纽约州州立大学水牛城分校，获数学博士学位。研究方向：随机过程、微分方程及其在生存分析中的应用、数理统计、空间计量、Topological data analysis以及算法设计。
国家发展研究院官方微信
Copyright© 1994-2012 北京大学 国家发展研究院 版权所有, 京ICP备05065075号-1
保留所有权利，不经允许请勿挪用
-
首页
-
教学项目展开 / 收起
sidenav background教学项目
sidenav header background[12月19日]发展经济学workshop
发布日期：2018-12-17 09:32 来源：北京大学国家发展研究院
时间：2018年12月19日（周三）14：00-15：30
地点：北大国发院/中国经济研究中心万众楼小教室
主持人：张晓波、李力行、席天扬
主讲人：张晓奇题目：Constrained clustering and two applications: detecting the impact of urban structure on the relationship between subway system and housing price & modeling the intra-urban housing market spillover
摘要：We develop a constrained clustering technique that has quite promising applications to topics in urban economics and housing market studies. I am going to present two applications of this method with different types of constraint.
The first application refers to detect the influence of urban structure on the relationship between subway system and housing prices, the following is the abstract:
Using housing market data of Beijing and Hangzhou, China, we conduct a case study to detect how the difference of urban structure can affect the relationship between subway system and housing prices. To quantify the characteristics of urban structure, we propose a constrained clustering method, which cannot only reveal the spatial heterogeneity of housing market, but also provides a link between the heterogeneity and the underlying urban structure. Applying constrained clustering to Beijing and Hangzhou, we find that the relationship between accessibility to metro stations and housing prices is weak and vulnerable, while the improvement of commuting efficiency, measured by a key variable – metro index, does have a robust connection to the metro premium on housing units. In particular, only a large metro index can associate with positive metro premium. Structural features, like the size of city core and the existence of multi-centres, influence the metro premium through affecting the value and spatial distribution of metro index. The evidence from Beijing and Hangzhou supports that in a mono-centric city, the size of city core is positively associated with metro index and metro premium, while in a polycentric city with a small city core, metro index tends to be lower in the core region and higher in the satellite regions, which enforces the metro premium to be negative in core while positive out core.The second application refers to modelling the intra-urban housing market spillover, the following is the abstract:
This study analyses the spillovers among intra-urban housing sub-markets in Beijing, China. Intra-urban spillover is at least as important as the widely-studied inter-urban spillover, because it is crucial to the inequality issue within a city which is always a main concern of city managers. Intra-urban spillover imposes methodological challenge. Unlike the inter-urban situation, the range of every submarket is not naturally defined, therefore it is not possible to directly evaluate intra-urban spillover by standard time-series models. A novel clustering technique, namely constrained clustering, is proposed to solve that issue. Using constrained clustering, spillover effect is formulated as a spatial Markov chain model, identification of submarkets as hidden states of Markov chain and estimation of the Markov transition matrix can be done simultaneously. Using a sampled day-by-day transaction dataset of second-hand apartments in Beijing from 2011 to 2017, constrained clustering detects that there exist quite robustly 16 submarkets within the urban-suburban area of Beijing, spillover effect transits among the 16 regions with the highest transition probability appearing in the overlapped region of urban core and Tongzhou district. This observation agrees with the new urban planning proposal announced in early 2012 and being continuously effective during the entire data collection period. In addition to identifying the impact of policy, we also analyses a variety of spillover “types” through regression analysis of Markov transition matrix and trajectory simulation of the spatial Markov chain. Our “type” analysis shows that the “ripple” form of spillover is no longer dominant at the intra-urban level, although it is believed to be the dominant type at inter-urban level in literature. Other types of spillover, such as the spillover due to the existence of price depressed regions, which are not significant at inter-urban level, but do play central roles at intra-urban level. This observation reflects the complexity of spillover dynamics in the intra-urban situation and the distinct driving-forces behind intra- and inter-urban spillovers.主讲人简介：张晓奇，浙江财经大学金融学院，讲师。2017年6月毕业于纽约州州立大学水牛城分校，获数学博士学位。研究方向：随机过程、微分方程及其在生存分析中的应用、数理统计、空间计量、Topological data analysis以及算法设计。
访问学者申请
EMAIL
北京大学
ENGLISH
