Method for identifying regional market structure by fusing feature price model and complex network

Abstract

The invention discloses a method for identifying a regional market structure by fusing a feature price model and a complex network. The method comprises the following steps: collecting housing price data of county cities throughout the country; combining the feature price model and the complex network to establish an urban area connection model; counting convergence indexes based on a prefecture-level city scale and reconstructing a network; and calculating the network by adopting a louvain algorithm to obtain a nationwide prefecture-level city scale regional structure. Based on a feature price model and a complex network, the invention provides a method for measuring the market convergence contact degree between cities through goodness-of-fit and further mining a regional structure. On the basis of a connection network between district and county city real estate markets, statistics and network reconstruction are carried out on the scale of prefecture-level cities, the fineness and operability of analysis are considered from the regional scale to the scale of prefecture-level cities, and comprehensive identification of the national real estate market regional structure is realized by capturing the microcosmic-macroscopic market similarity.

Description

Technical Field

[0001] The present invention relates to the field of regional structure research, and in particular to a method for identifying regional market structure by integrating a characteristic price model and a complex network. Background Art

[0002] With the acceleration of urbanization and the continuous development of the real estate market, the study of regional market structure has become an important topic in urban planning and economic development. Urbanization has brought about the concentration of population and the reconfiguration of resources, which has led to closer economic ties between cities. As an important part of the urban economy, the structural changes of the real estate market directly affect the sustainable development of the city. Studying the regional market structure helps to reveal the potential mechanism of the market, identify the key factors affecting housing prices, and provide a scientific basis for policy making. In addition, the analysis of the regional market structure can also reflect the interactive relationship between cities, reveal the flow trend of economic activities within and outside the region, and help planners to rationally allocate resources and optimize the functional layout of the city.

[0003] Traditional regional division methods often rely on dimensions such as geographical location, administrative divisions, human traffic or patent industry associations. Although these methods reveal regional characteristics to a certain extent, they often cannot fully reflect the market correlation and interactive relationship between different regions when facing the complex and ever-changing real estate market. Therefore, in the current rapidly changing economic environment, relying solely on these traditional division methods is insufficient. The hedonic price model, based on consumers' preferences for different characteristics, reveals the various factors that affect urban market connections. Its advantage is that it can quantify complex market factors, thereby providing strong support for regional economic analysis. On the other hand, complex network theory can reveal the interactions between cities and their structural characteristics by treating cities as nodes in the network and market connections as edges.

[0004] Therefore, in order to more fully understand the complexity of regional markets, a regional structure identification method that takes into account market connections and interactions is urgently needed to achieve in-depth analysis of regional market structure. Models that integrate hedonic price models and complex networks can measure the complex connections between urban markets and then identify regional market structures. However, there is currently no complete method that integrates hedonic price models and complex network models to identify regional structures. Summary of the invention

[0005] In view of the above problems existing in the prior art, the present invention provides a method for identifying regional market structure by integrating the characteristic price model and complex network. Based on the similarity of the internal mechanism of the real estate market, the method combines the complex network and the characteristic price model to quantify the connection strength between urban markets and thus identify the regional market structure. Measuring the connection between urban real estate markets through the characteristic price model to establish a new understanding of the urban regional structure can provide a new perspective and method for understanding the regional market structure and policy implementation.

[0006] The technical solution of the present invention is as follows:

[0007] A method for identifying regional market structure by integrating a hedonic price model and a complex network comprises the following steps:

[0008] S1. Collect housing price data of cities and counties across the country, define them as housing price sample points, and obtain a set HP; the housing price sample points include: housing price information, housing structure information and housing latitude and longitude information; the housing structure information includes: area, number of bedrooms, number of bathrooms and floors;

[0009] S2. Combine the characteristic price model and complex network to establish a city-region connection model. The specific steps are as follows:

[0010] S2-1. Taking county-level cities as research units, we construct an initial network of cities with shared edges, denoted as network G0(V,E); each county-level city is a node in the network, called a county-level city node;

[0011] S2-2. Based on the housing price sample points in the set HP, a characteristic price model is established for county-level cities. The variables used include: housing structure information, location characteristics and facility distance; the location characteristics refer to the road selectivity and integration calculated using the spatial syntax software Depthmap; the facility distance refers to the distance between the house and the hospital, primary school, middle school, and supermarket;

[0012] S2-3. Select any two adjacent district, county, and city nodes i and j in the G0(V, E) network and calculate the goodness of fit of their respective characteristic price models, denoted as R i 2 With R j 2 ; Combine the data of i and j, and calculate the goodness of fit of the combined characteristic price model again, denoted as R ij 2 ;

[0013] S2-4. Calculate the network edge weight W1, that is:

[0014]

[0015] S2-5, assign weights to the G0(V,E) network; V is the district, county, or city node, and the node weight is the goodness of fit of the district, county, or city node; E is the network edge between district, county, or city nodes i and j, and the weight of the network edge is W1;

[0016] S2-6. Remove the edges with W1 less than 1 in G0(V,E), and obtain the regional connection network G1(V,E) at the county scale;

[0017] S3. Statistical convergence indicators based on prefecture-level cities and reconstruct the network. The specific steps are as follows:

[0018] S3-1. Taking prefecture-level cities as research units, we construct an initial network of cities with shared edges, denoted as network G2(V,E); each prefecture-level city is a node in the network, called a prefecture-level city node;

[0019] S3-2. Count the number of edges in the G1(V,E) network within the prefecture-level city as the node weight of the prefecture-level city node in the G2(V,E) network;

[0020] S3-3. Count the number of edges in the G1(V,E) network between any adjacent prefecture-level cities, and use it as the edge weight W2 in the G2(V,E) network;

[0021] S3-4. Remove the edges with W2=0 in G2(V,E) to obtain the regional convergence network G3(V,E) at the prefecture-level city scale;

[0022] S4. Use the Louvain algorithm to calculate G3(V,E) to obtain the regional structure of prefecture-level cities across the country. The specific steps are as follows:

[0023] S4-1, initialize the maximum modularity and the best resolution, select the resolution range and step size, and divide the resolution range into a series of resolutions with the step size as the unit;

[0024] S4-2, repeatedly execute the Louvain algorithm for each resolution and record the community division results each time; select the value of the random seed;

[0025] S4-3. Count the number of times each prefecture-level city node is assigned to different communities, generate community assignment probabilities, and select the community to which the prefecture-level city node is most frequently assigned as the final division;

[0026] S4-4. Calculate the modularity of the division, and finally output the maximum modularity and its corresponding optimal resolution, so as to obtain the regional structure of the prefecture-level city scale across the country.

[0027] Furthermore, in step S4-1, the maximum modularity and the best resolution are -1, the step size is 0.1, and the resolution range is 0.1 to 2.0;

[0028] In step S4-1, the maximum modularity and the best resolution are -1, the step size is 0.1, and the resolution range is 0.1 to 2.0;

[0029] In step S4-2, the Louvain algorithm is repeatedly executed 10 times; the value of the random seed is 42.

[0030] The beneficial technical effects of the present invention are:

[0031] (2) Based on the characteristic price model and complex network, this paper proposes a method to measure the degree of market convergence between cities by goodness of fit, and then to mine the regional structure;

[0032] (3) The regional structure identification method proposed in the present invention is based on the connection network between the real estate markets of counties and cities, and then statistically analyzes and reconstructs the network at the prefecture-level city scale. The purpose of moving from the regional scale to the prefecture-level city scale is to balance the precision and operability of the analysis. By capturing the market similarities from micro to macro, a comprehensive identification of the regional structure of the national real estate market can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic diagram of the steps of the present invention;

[0034] Figure 2 It is the national regional structure identified based on the present invention. DETAILED DESCRIPTION

[0035] The present invention is described in detail below in conjunction with the accompanying drawings and embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] The embodiment adopts the present invention to identify the regional structure of cities across China. The specific steps are as follows:

[0037] S1. Collect housing price data of prefecture-level cities in China by crawling second-hand housing transaction websites such as Anjuke and Beike (or other data acquisition methods), including house unit price information, house structural attributes such as area, number of bedrooms, number of bathrooms, and floors to facilitate the subsequent establishment of a characteristic price model, and longitude and latitude information for geocoding processing into shapefile spatial data.

[0038] The housing price data of each prefecture-level city was corrected to obtain data in the WGS1984 coordinate system, and the spatial location filtering tool in ArcGIS Pro was used to filter the housing price data of each district and county using China's district and county-level administrative division data (also in the WGS1984 coordinate system). Among them, there are 1,515 district and county cities with more than 100 housing price data.

[0039] S2. Combine the characteristic price model and complex network to establish a city-region connection model. The specific steps are as follows:

[0040] S2-1. Call the geopandas library in Python to use the intersects function to calculate the adjacency matrix of the shp file at the county level across the country, and then call the undirected graph function in the networkx library in Python to build an initial network of city connections with common edges, recorded as network G0(V,E) (such as Figure 1 )

[0041] S2-2. For cities and districts with more than 100 housing price points, a characteristic price model is established. The variables used include: house structure information, location characteristics and facility distance. House structure information includes area, number of bedrooms, number of bathrooms, floors, etc. in data collection. Location characteristics include road selectivity and integration calculated using the spatial syntax software Depthmap. Facility distance includes the distance to hospitals, primary schools, middle schools, supermarkets and other facilities.

[0042] S2-3. Call the statsmodels library in Python to use the variance_inflation_factor function to calculate multicollinearity for all variables, and gradually eliminate the factors with the highest VIF (variance inflation factor) and calculate VIF again until the VIF of all variables is less than or equal to 5, and obtain the variables for constructing the final model.

[0043] S2-4. Select any two adjacent nodes i and j in the G0(V,E) network, call the statsmodels library in Python to use the OLS function to perform multiple linear regression, and obtain their respective goodness of fit, denoted as R i 2 With R j 2 ; Merge the data of i and j, call the statsmodels library in Python again to use the OLS function to perform multiple linear regression, and obtain the goodness of fit of the characteristic price model after the merged data, recorded as R ij 2 ;

[0044] S2-5. Calculate the network edge weight W1, that is:

[0045]

[0046] S2-6. Assign weights to the G0(V,E) network. In the network G0(V,E), V is a city node, the node weight is the goodness of fit of the city node, E is a network edge, and the weight of the edge is W1;

[0047] S2-7. Remove the edges with W1 less than 1 in the network G0(V,E), and obtain the regional connection network G1(V,E) at the county scale (e.g. Figure 1 )

[0048] S3. Statistical convergence indicators based on prefecture-level cities and reconstruct the network. The specific steps are as follows:

[0049] S3-1. Call the geopandas library in Python to use the intersects function to calculate the adjacency matrix of the shp file at the prefecture-level city level across the country, and then call the undirected graph function in the networkx library in Python to build an initial network of prefecture-level cities with shared edges, recorded as network G2(V,E) (such as Figure 1 )

[0050] S3-2, count the number of edges in the G1 network within the prefecture-level city as the node weight of the G2 network;

[0051] S3-3. Count the number of edges in the G1 network between any adjacent prefecture-level cities, and use it as the edge weight W2 in the G2 network;

[0052] S3-4, delete the edges with W2=0 in the G2 network, and obtain the regional convergence network G3(V,E) at the prefecture-level city scale;

[0053] S4. Call the Louvain algorithm in the community function in Python to calculate the regional convergence network G3(V,E) at the prefecture-level city scale, initialize the maximum modularity and the best resolution to -1, and then run the Louvain algorithm multiple times in the resolution range of 0.1 to 2.0 (step size of 0.1) to find the best community division. For each resolution, the Louvain algorithm will run 10 times, and the community division results of each time will be recorded. The random seed is selected as 42. Then count the number of times each node is assigned to different communities, generate the community assignment probability, and select the community to which the node is most frequently assigned as the final division. Subsequently, calculate the modularity of the division (used to evaluate the quality of community division), and finally output the maximum modularity and its corresponding best resolution, and then obtain the national prefecture-level city-scale regional structure, such as Figure 1 As shown in the lower right figure.

[0054] S5. Call the GeoPandas and Matplotlib libraries in Python to visualize the regional market structure.

[0055] like Figure 2 As shown in the figure, based on the existing data, the regional market structure obtained by this method identified 34 communities, of which 15 communities had a scale (number of nodes) greater than or equal to 10. This example shows the regional structure division based on real estate market connections across the country (with existing data).

[0056] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and implementation modes, and can be fully applicable to various fields suitable for the present invention. For those familiar with the art, for those of ordinary skill in the art, various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to specific details.

Claims

1. A method for identifying regional market structure by integrating hedonic price model and complex network, characterized in that: The following steps are involved: S1. Collect housing price data of cities and counties across the country, define them as housing price sample points, and obtain a set HP; the housing price sample points include: housing price information, housing structure information and housing latitude and longitude information; the housing structure information includes: area, number of bedrooms, number of bathrooms and floors; S2. Combine the characteristic price model and complex network to establish a city-region connection model. The specific steps are as follows: S2-1. Taking county-level cities as research units, we construct an initial network of cities with shared edges, denoted as network G0(V,E); each county-level city is a node in the network, called a county-level city node; S2-2. Based on the housing price sample points in the set HP, a characteristic price model is established for county-level cities. The variables used include: housing structure information, location characteristics and facility distance; the location characteristics refer to the road selectivity and integration calculated using the spatial syntax software Depthmap; the facility distance refers to the distance between the house and the hospital, primary school, middle school, and supermarket; S2-3. Select any two adjacent district, county, and city nodes i and j in the G0(V, E) network and calculate the goodness of fit of their respective characteristic price models, denoted as R i 2 With R j 2 ; Combine the data of i and j, and calculate the goodness of fit of the combined characteristic price model again, denoted as R ij 2 ; S2-4. Calculate the network edge weight W1, that is: S2-5, assign weights to the G0(V,E) network; V is the district, county, or city node, and the node weight is the goodness of fit of the district, county, or city node; E is the network edge between district, county, or city nodes i and j, and the weight of the network edge is W1; S2-6. Remove the edges with W1 less than 1 in G0(V,E), and obtain the regional connection network G1(V,E) at the county scale; S3. Statistical convergence indicators based on prefecture-level cities and reconstruct the network. The specific steps are as follows: S3-1. Taking prefecture-level cities as research units, we construct an initial network of cities with shared edges, denoted as network G2(V,E); each prefecture-level city is a node in the network, called a prefecture-level city node; S3-2. Count the number of edges in the G1(V,E) network within the prefecture-level city as the node weight of the prefecture-level city node in the G2(V,E) network; S3-3. Count the number of edges in the G1(V,E) network between any adjacent prefecture-level cities, and use it as the edge weight W2 in the G2(V,E) network; S3-4. Remove the edges with W2=0 in G2(V,E) to obtain the regional convergence network G3(V,E) at the prefecture-level city scale; S4. Use the Louvain algorithm to calculate G3(V,E) to obtain the regional structure of prefecture-level cities across the country. The specific steps are as follows: S4-1, initialize the maximum modularity and the best resolution, select the resolution range and step size, and divide the resolution range into a series of resolutions with the step size as the unit; S4-2, repeatedly execute the Louvain algorithm for each resolution and record the community division results each time; select the value of the random seed; S4-3. Count the number of times each prefecture-level city node is assigned to different communities, generate community assignment probabilities, and select the community to which the prefecture-level city node is most frequently assigned as the final division; S4-4. Calculate the modularity of the division, and finally output the maximum modularity and its corresponding optimal resolution, so as to obtain the regional structure at the prefecture-level city level across the country.

2. The method for identifying regional market structure by integrating the characteristic price model and complex network according to claim 1, characterized in that: In step S4-1, the maximum modularity and the best resolution are -1, the step size is 0.1, and the resolution range is 0.1 to 2.0; In step S4-2, the Louvain algorithm is repeatedly executed 10 times; the value of the random seed is 42.

Images