WO2022227219A1 - 房产指数数据分析方法、装置、设备及存储介质 - Google Patents

房产指数数据分析方法、装置、设备及存储介质 Download PDF

Info

Publication number
WO2022227219A1
WO2022227219A1 PCT/CN2021/097416 CN2021097416W WO2022227219A1 WO 2022227219 A1 WO2022227219 A1 WO 2022227219A1 CN 2021097416 W CN2021097416 W CN 2021097416W WO 2022227219 A1 WO2022227219 A1 WO 2022227219A1
Authority
WO
WIPO (PCT)
Prior art keywords
time series
series data
factor
target
target time
Prior art date
Application number
PCT/CN2021/097416
Other languages
English (en)
French (fr)
Inventor
段洪云
Original Assignee
平安科技(深圳)有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 平安科技(深圳)有限公司 filed Critical 平安科技(深圳)有限公司
Publication of WO2022227219A1 publication Critical patent/WO2022227219A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/06Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063Operations research, analysis or management
    • G06Q10/0637Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
    • G06Q10/06375Prediction of business process outcome or impact based on a proposed change
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24Querying
    • G06F16/245Query processing
    • G06F16/2458Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2474Sequence data queries, e.g. querying versioned data
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10Services
    • G06Q50/16Real estate
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10Services
    • G06Q50/26Government or public services

Definitions

  • the present application relates to the technical field of data analysis, and in particular, to a method, device, device and storage medium for analyzing real estate index data.
  • the real estate industry refers to the industry engaged in the development, operation, management and service of land and real estate.
  • external environmental factors such as Consumer Price Index (CPI), Producer Price Index (PPI), steel sales, export of home appliances, housing provident fund loan rate, construction area and
  • CPI Consumer Price Index
  • PPI Producer Price Index
  • steel sales export of home appliances
  • housing provident fund loan rate
  • construction area and
  • These external environmental factors are closely related to the real estate index, which is in line with the real estate industry factors, such as the sales price and sales area of real estate.
  • the inventor realizes that most of the current real estate industry uses statistical methods to predict the correlation between external environmental factors and real estate indices, such as calculating the correlation prediction of external environmental factors on real estate indices by cyclically setting the lag order.
  • the real estate index does not conform to the laws of the real estate industry and the macroeconomic cycle, resulting in inaccurate predictions of the real estate index by external environmental factors, which is not conducive to the government's accurate and effective implementation of real estate regulation, and it is difficult to ensure the accuracy of subsequent real estate industry expectations.
  • the main purpose of the present application is to provide a real estate index data analysis method, device, equipment and storage medium, aiming at improving the accuracy of real estate index analysis and prediction, thereby facilitating the accurate and effective implementation of real estate related regulation.
  • the present application provides a real estate index data analysis method, including:
  • first time series data of a first factor and second time series data of a second factor, where the first factor is a real estate index factor conforming to the real estate industry element, and the second factor is an external environmental factor affecting the development of the real estate industry;
  • the present application further provides a real estate index data analysis device, the real estate index data analysis device comprising:
  • the acquisition module is used to acquire the first time series data of the first factor and the second time series data of the second factor, where the first factor is a real estate index factor conforming to the real estate industry element, and the second factor is a factor that affects the development of the real estate industry external environmental factors;
  • a processing module configured to perform seasonal adjustment and smoothing processing on both the first time series data and the second time series data to obtain the first target time series data and the second target time series data;
  • a first determination module configured to calculate the transfer entropy between the first target time series data and the second target time series data, and determine the first leading indicator between the first factor and the second factor based on the transfer entropy information
  • a second determining module configured to determine a linear relationship between the first target time series data and the second target time series data, and determine a second leading indicator between the first factor and the second factor based on the linear relationship information
  • a third determining module configured to determine target leading indicator information between the first factor and the second factor according to the first leading indicator information and the second leading indicator information;
  • a prediction module configured to obtain the current time series data of the second factor, and predict the target time series data of the first factor based on the target leading indicator information and the current time series data.
  • the present application also provides a computer device, the computer device comprising a processor, a memory, and a computer program stored on the memory and executable by the processor, wherein the computer program is executed by the When the processor executes, it implements the following steps:
  • first time series data of a first factor and second time series data of a second factor, where the first factor is a real estate index factor conforming to the real estate industry element, and the second factor is an external environmental factor affecting the development of the real estate industry;
  • the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, wherein when the computer program is executed by the processor, the following steps are implemented:
  • first time series data of a first factor and second time series data of a second factor, where the first factor is a real estate index factor conforming to the real estate industry element, and the second factor is an external environmental factor affecting the development of the real estate industry;
  • the present application provides a real estate index data analysis method, device, equipment and storage medium.
  • the present application obtains first time series data of a first factor and second time series data of a second factor, where the first factor is a property that conforms to the real estate industry elements Exponential factor; perform seasonal adjustment and smoothing processing on both the first time series data and the second time series data to obtain the first target time series data and the second target time series data; calculate the difference between the first target time series data and the second target time series data and determine the first leading indicator information between the first factor and the second factor based on the transfer entropy; determine the linear relationship between the first target time series data and the second target time series data, and determine the first The second leading indicator information between the factor and the second factor; according to the first leading indicator information and the second leading indicator information, determine the target leading indicator information between the first factor and the second factor; obtain the current time series of the second factor data, and predict the target time series data of the first factor based on the target leading indicator information and the current time series data.
  • FIG. 1 is a schematic flowchart of steps of a method for analyzing real estate index data provided by an embodiment of the present application
  • Fig. 2 is a schematic flow chart of a sub-step of the real estate index data analysis method in Fig. 1;
  • Fig. 3 is another sub-step flowchart schematic diagram of the real estate index data analysis method in Fig. 1;
  • FIG. 4 is a schematic block diagram of a real estate index data analysis device provided by an embodiment of the present application.
  • FIG. 5 is a schematic block diagram of a sub-module of the real estate index data analysis device in FIG. 4;
  • FIG. 6 is a schematic block diagram of another sub-module of the real estate index data analysis device in FIG. 4;
  • FIG. 7 is a schematic structural block diagram of a computer device according to an embodiment of the present application.
  • Embodiments of the present application provide a method, apparatus, device, and storage medium for analyzing real estate index data.
  • the real estate index data analysis method can be applied to a terminal device or a server, and the terminal device can be an electronic device such as a mobile phone, a tablet computer, a notebook computer, a desktop computer, a personal digital assistant and a wearable device; the server can be a single It can also be a server cluster consisting of multiple servers.
  • the following takes the real estate index data analysis method applied to the server as an example for explanation.
  • FIG. 1 is a schematic flowchart of steps of a method for analyzing real estate index data according to an embodiment of the present application.
  • the real estate index data analysis method includes steps S101 to S106.
  • Step S101 acquiring first time series data of the first factor and second time series data of the second factor.
  • the first factor is a real estate index factor that conforms to the elements of the real estate industry, and the first factor can be a low-frequency factor that has economic cycle fluctuations and is related to the real estate industry in a macro sense.
  • the first factor is the sales price of commercial housing in various regions. , sales area or sales volume growth rate, and for example, the first factor is the second-hand housing transaction price, transaction volume or transaction growth in each region.
  • the second factor is an external environmental factor that affects the development of the real estate industry.
  • the second factor is Consumer Price Index (CPI), Producer Price Index (PPI), steel sales, home appliance exports, housing provident fund loans rate, construction area or de-chemical cycle, etc.
  • CPI Consumer Price Index
  • PPI Producer Price Index
  • the first time series data of the first factor may be a data column uniformly recorded by the same real estate index factor in chronological order, and each data in the same data column is of the same caliber and is comparable.
  • the first time series data may be the number of epochs or the number of time points.
  • the second time series data of the second factor may be a data column uniformly recorded in chronological order by the same external environmental factor, and may be the number of periods or points in time, which is not specifically limited in this embodiment.
  • the server obtains the first time series data of the first factor and the second time series data of the second factor from the memory or the cloud, or obtains the first time series data of the first factor and the second time series data from different memories or databases respectively.
  • the second time series data of the second factor for example, the first time series data of the first factor is obtained from the first database, and the second time series data of the second factor is obtained from the second database, which is not specifically limited in this embodiment. .
  • the related information such as the first time series data and the second time series data can also be stored in a blockchain node.
  • the technical solution of this application can also be applied to add other data files stored on the blockchain, and the blockchain referred to in this application is a new application of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, encryption algorithm, etc. model.
  • Blockchain essentially a decentralized database, is a series of data blocks associated with cryptographic methods. Each data block contains a batch of network transaction information to verify its Validity of information (anti-counterfeiting) and generation of the next block.
  • the blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.
  • Step S102 Perform seasonal adjustment and smoothing processing on both the first time series data and the second time series data to obtain the first target time series data and the second target time series data.
  • the processing dimension of the time series data in this embodiment is more comprehensive.
  • the influence of implicit seasonal factors in the first time series data and the second time series data is corrected, so that the first time series data and the second time series data in different seasons are comparable , so that the adjusted first target time series data and the second target time series data can better reflect the actual change trend.
  • the economic significance of the first time series data and the second time series data can be better extracted, so that the obtained first target time series data and the second target time series data are more Stable, to further ensure the validity of the data and the accuracy of subsequent real estate index analysis and prediction.
  • step S102 includes: sub-steps S1021 to S1022 .
  • Sub-step S1021 respectively process the first time series data and the second time series data by using a preset ARIMA model to obtain the adjusted first time series data and the second time series data.
  • the ARIMA model includes a reg ARIMA model or a seasonal ARIMA model that considers calendars, holidays, and trading days.
  • the ARIMA model is used to preprocess the first time series data and the second time series data, respectively, to correct the influence of the hidden seasonal factors in the first time series data and the second time series data, so that the time series data of different seasons are comparable, so that the adjusted time series data can be compared.
  • the first time series data and the second time series data can better reflect the actual change trend, thereby improving the accuracy of the real estate index analysis and prediction.
  • Sub-step S1022 respectively process the adjusted first time series data and the second time series data by using a preset empirical modal decomposition EMD model to obtain the first target time series data and the second target time series data.
  • the empirical mode decomposition EMD model performs signal decomposition according to the time scale characteristics of the adjusted first time series data and the second time series data itself, and does not need to set any basis function in advance.
  • the empirical mode decomposition EMD model performs signal decomposition according to the time scale characteristics of the adjusted first time series data and the second time series data itself, and does not need to set any basis function in advance.
  • the method further includes: respectively adjusting the first target time series data and the second time series data.
  • the target time series data is filtered to obtain the filtered first time series data and the filtered second target time series; the filtered first time series data is used as the updated first target time series data, and the filtered second target time series is used as the updated first time series. 2.
  • Target time series data It should be noted that by filtering the first target time series data and the second target time series data respectively, the noise in the first target time series data and the second target time series data can be prevented from negatively affecting subsequent modeling.
  • the filtering method adopted in this embodiment filters the noises in the first target time series data and the second target time series data respectively, eliminates the influence of invalid and abnormal data, and further ensures the validity of the data and the accuracy of the analysis and prediction of the real estate index.
  • performing seasonal adjustment and smoothing processing on both the first time series data and the second time series data, and after obtaining the first target time series data and the second target time series data further includes: performing the second target time series data on the second target time series data. Stationarity test to obtain a test result; if the test result is that the second target time series data is stable, then execute the step of calculating the transfer entropy between the first target time series data and the second target time series data; if the test result is the second target time series data If it is not stable, the target leading indicator information between the first factor and the second factor is determined by the preset Barzilai-Borwein gradient algorithm.
  • the unit root ADF test method can be used to test the stationarity of the target time series data. For example, assuming that the second target time series data is stable, and calculate the p value. If the p value is less than 0.05, the null hypothesis is rejected. It should be noted that, after seasonal adjustment and stabilization processing, the second target time series data is not necessarily stable.
  • the first factor is calculated by performing a stationarity test on the second target time series data and selecting different implementation strategies according to the test results.
  • the target leading indicator information between the second factor and the target leading indicator information is more accurate, which is conducive to further improving the accuracy of the real estate index analysis and prediction.
  • determining the second leading indicator information between the first factor and the second factor includes: respectively performing piecewise linear processing on the first target time series data and the second target time series data , to determine the first peak and the first trough of the first target time series data, and to determine the second peak and the second trough of the second target time series data; determine the target peak before the first trough from the plurality of second peaks, and calculate the time difference between the target wave crest and the first wave trough; and determine the target wave trough before the first wave crest from a plurality of second wave troughs, and calculate the time difference between the target wave trough and the first wave crest; according to the target wave The time difference between the peak and the first trough, and the time difference between the target trough and the first peak, determine the first lead period of the second factor with respect to the first factor.
  • an upward or downward trend of the first target time series data and the second target time series data can be obtained.
  • first target time series data as an example, after performing piecewise linear processing on the first target time series data, multiple extreme points (including maximum points and minimum points) are obtained, and the iteration starts from the first extreme point. , the two adjacent extreme points are required to be in opposite directions, that is, if the first one is a peak, the next linearized endpoint must be a trough. The first peak and first trough of the target time series data. Similarly, the second peak and the second trough of the second target time series data can also be obtained by piecewise linear processing.
  • the first wave trough and the second wave peak are matched, and the target wave peak before the first wave trough is determined from the plurality of second wave peaks, that is, the target wave peak in the plurality of second wave peaks before the first wave trough is selected.
  • the target peak before the preset time range of the first trough can also be determined from a plurality of second peaks, and the preset time range can be set flexibly.
  • the preset time range is 6 months, that is, the peak before the first trough is selected.
  • a target wave trough preceding the first wave crest can be determined from the plurality of second wave troughs in the same manner.
  • the value obtained by dividing the number of matched target peaks by the total number of second troughs is considered to be the matching probability with the first target time series data.
  • the probability and preset threshold determine whether the second factor has a leading effect on the first factor.
  • the preset threshold is 0.7. If the matching probability is lower than 0.7, it is considered that the two sequences do not match, that is, the second factor does not have a leading role; if the matching probability is greater than or equal to 0.7, it is considered that the first target time series data and the second target time series data The data are matched, at which point the second factor is considered to have a leading effect on the first factor.
  • the first lead period of the second factor to the first factor is determined by calculating the average value of the time difference between the target peak and the first trough and the time difference between the target trough and the first peak.
  • a first lead period of the second factor on the first factor is determined.
  • the first lead period is the average value of the difference between the peak and trough times that match each other in the first target time series data and the second target time series data.
  • determining the first lead period of the second factor for the first factor includes: calculating the target peak Calculate the time difference between the target wave trough and each first wave trough to obtain multiple first time difference values; calculate the time difference value between the target wave trough and each first wave peak to obtain multiple second time difference values; calculate multiple The average value of the first time difference value and the plurality of second time difference values is used to obtain the first lead period of the second factor to the first factor.
  • Step S103 Calculate the transfer entropy between the first target time series data and the second target time series data, and determine the first leading indicator information between the first factor and the second factor based on the transfer entropy.
  • the first leading indicator information includes a lead-lag relationship between the first factor and the second factor, such as the lead period or the lag period of the second factor to the first factor. It should be noted that determining the first leading indicator information between the first factor and the second factor based on the transfer entropy facilitates capturing the nonlinear leading relationship between the first target time series data and the second target time series data.
  • the concept of one-way transfer entropy is applied to time series data, which is conducive to constructing a net transfer ratio to eliminate the influence of dimensions, thereby improving the accuracy of real estate index analysis and prediction.
  • step S103 includes: sub-steps S1031 to S1034 .
  • Sub-step S1031 Calculate the change amount of the uncertainty of the first target time series data to the second target time series data to obtain the first transfer entropy.
  • the calculation formula of transfer entropy can be: p and q represent the two data distributions, respectively.
  • the transfer entropy represents the change of the uncertainty of the information of y to the information of x.
  • This embodiment innovatively applies the concept of transfer entropy to the time series to capture the nonlinear relationship between the two target time series data. lead relationship. Therefore, if the first target time series data is denoted as Xn, and the second target time series data is denoted as Yn, the first transfer entropy for Yn at time Xn-t is denoted as H(x->y), where t is time measurement.
  • Sub-step S1032 Calculate the change amount of the uncertainty of the second target time series data with respect to the first target time series data to obtain the second transfer entropy.
  • the second target time series data is recorded as Yn
  • the information transfer from Yn-t time to Xn is calculated at the same time
  • the second transfer entropy is recorded as H(y->x) .
  • Sub-step S1033 Calculate the target transfer entropy between the first target time series data and the second target time series data according to the first transfer entropy and the second transfer entropy.
  • the target transfer entropy is calculated.
  • a lead-lag relationship between the second factor and the first factor is determined according to the first transfer entropy and the second transfer entropy. Calculate the difference H(x->y)-H(y->x) between the first transfer entropy and the second transfer entropy. If the difference is positive, it means that the second factor has a lead over the first factor If the difference is negative, it means that the second factor has a hysteresis relationship with the first factor.
  • the calculation formula of target transfer entropy is:
  • the target transfer entropy is TE
  • the first transfer entropy is H1
  • the second transfer entropy is H2.
  • can be set to 0.00001
  • abs(H1) is the absolute value of the first transfer entropy H1
  • abs(H2) is the absolute value of the second transfer entropy H2.
  • Sub-step S1034 based on the target transfer entropy, determine the first lead period of the second factor for the first factor.
  • the value range of the target transfer entropy is (0, 1)
  • the second factor affects the first value according to the target transfer entropy and the preset net transfer ratio.
  • factor has a leading role.
  • the default net transfer ratio is 0.5. If the target transfer entropy is lower than the preset net transfer ratio, the second factor has no leading effect; if the target transfer entropy is greater than or equal to the preset net transfer ratio, the second factor has a leading effect on the first factor.
  • the first transfer entropy H(x->y) includes the information transfer to Yn at time Xn-t, then there are multiple first transfer entropies, and the second transfer entropy H(y->x) includes Yn -In the information transfer from time t to Xn, there are multiple second transfer entropies, where t is a time measurement, such as day, week, month, or year.
  • t is a time measurement, such as day, week, month, or year.
  • target transfer entropies Select the first target transfer entropy or the second target transfer entropy corresponding to the largest target transfer entropy, and determine the time series difference corresponding to the first target transfer entropy or the second target transfer entropy, and obtain the second factor for the first factor.
  • a lead period wherein the time series difference is the time difference between the first time series data of the first factor and the second time series data of the second factor.
  • the time metric t is a month
  • the first time series data includes a time series of real estate index factors from January to December 2020
  • the second time series data includes a time series of external environmental factors from January to December 2020.
  • the maximum target transfer entropy is 0.8
  • the corresponding first target transfer entropy is calculated using the first time series data in June 2020 and the second time series data in October 2020, that is, the time corresponding to the first target transfer entropy
  • the sequence difference is the time difference between October 2020 and June 2020
  • the time difference between the external environmental factor and the real estate index factor is 4 months, that is, the first leading period of the second factor to the first factor is 4 months.
  • Step S104 Determine a linear relationship between the first target time series data and the second target time series data, and determine second leading indicator information between the first factor and the second factor based on the linear relationship.
  • the second leading indicator information includes a leading-lag relationship between the first factor and the second factor, such as a lead period or a lag period of the second factor to the first factor. It should be noted that, because the aforementioned calculation of the transfer entropy between the first target time series data and the second target time series data discretizes the data, the capture of the linear relationship is slightly weaker than the linear correlation, and the output result of the first leading indicator information is Biased to include more nonlinear correlation factors.
  • determining the linear relationship between the first target time series data and the second target time series data, and determining the second leading indicator information between the first factor and the second factor based on the linear relationship is conducive to comprehensively integrating the first target time series data
  • the linear relationship and nonlinear relationship between the second target time series data, improve the accuracy of real estate index analysis and prediction.
  • the causal relationship between the first target time series data and the second target time series data is calculated, and the second factor for the first factor is determined based on the causal relationship. lead period.
  • the accuracy of the historical value of the first target time series data and the second target time series data for the predicted value is studied through the Granger causality test algorithm, and the statistics between the first target time series data and the second target time series data are obtained. Learn causality, and determine the second factor based on statistical causality.
  • the nonlinear correlation between the first target time series data and the second target time series data is supplemented, which can improve the real estate index. Analysis and forecasting accuracy.
  • the regression formula of Granger causality test is: Among them, u 1t is white noise, ⁇ and ⁇ are set coefficients, y is the first target time series data of the first factor, and x is the second target time series data of the second factor.
  • a plurality of lag term coefficients between the first target time series data and the second target time series data are calculated, and based on the plurality of lag term coefficients, the second factor is determined for the first time series.
  • a factor of the second lead period can be selected from the multiple lag term coefficients as the lead period of the second factor for the first factor, or the average value of the multiple lag term coefficients can be calculated, and the average value of the multiple lag term coefficients can be calculated. The average is taken as the lead period.
  • rl is the lag term coefficient
  • y is the first target time series data of the first factor
  • x is the second target time series data of the second factor.
  • the first target time series data and the second target time series data are processed based on a preset Granger causality test algorithm and a lag correlation coefficient algorithm to determine the second lead period of the second factor to the first factor.
  • a preset Granger causality test algorithm and a lag correlation coefficient algorithm to determine the second lead period of the second factor to the first factor.
  • other linear correlation algorithms or models in the art may also be used to calculate the second leading indicator information between the first factor and the second factor, which is not specifically limited in this embodiment.
  • Step S105 Determine target leading indicator information between the first factor and the second factor according to the first leading indicator information and the second leading indicator information.
  • the target leading indicator information includes a lead-lag relationship between the first factor and the second factor, for example, the lead period or the lag period of the second factor to the first factor.
  • first leading indicator information tends to contain more nonlinear correlation factors, and captures the nonlinear relationship strongly.
  • second leading indicator information emphasizes the capture of linearly related factors, and tends to screen out many linearly related factors.
  • the ultimate goal of this application is to screen out factors that have both linear and nonlinear relationships. Therefore, the target leading indicator information can be determined according to the first leading indicator information and the second leading indicator information, which can greatly improve the accuracy of real estate index analysis and prediction.
  • the intersection of the first leading indicator information and the second leading indicator information is obtained to obtain target leading indicator information between the first factor and the second factor.
  • the first leading indicator information is that the first leading period of the second factor to the first factor is 4 months
  • the second leading indicator information is that the second leading period of the second factor to the first factor is 2 months. If the first leading indicator information and the second leading indicator information are intersected, the target leading indicator information is that the leading period of the second factor to the first factor is 2 months to 4 months.
  • both the first leading indicator information and the second leading indicator information include a time range of a lead-lag relationship between the second factor and the first factor, and according to the first leading indicator information and the second leading indicator information, The lead term of the second factor over the first factor is determined.
  • the first leading indicator information is that the first leading period of the second factor to the first factor is 1-4 months
  • the second leading indicator information is that the second leading period of the second factor to the first factor is 0-2 months month
  • the target leading indicator information is that the leading period of the second factor to the first factor is 1-2 months.
  • the calculation methods include linear correlation algorithms such as calculation transfer entropy algorithm, cross entropy algorithm, Granger causality test algorithm, lag correlation algorithm, etc.
  • linear correlation algorithms such as coefficient algorithm
  • coefficient algorithm the smaller the common intersection of the first leading indicator information and the second leading indicator information, the larger the influence of noise, and the smaller the range of the target leading indicator period.
  • Step S106 Obtain the current time series data of the second factor, and predict the target time series data of the first factor based on the target leading indicator information and the current time series data.
  • the target leading indicator information includes the lead period of the second factor to the first factor. Based on the lead period of the second factor to the first factor and the current time series data of the second factor, it can be predicted that the second factor will lead to the first factor Target time series data for the first factor after the factor's lead period. Therefore, real estate regulation can be implemented in a targeted manner based on the target time series data to ensure that the real estate industry can develop as expected.
  • the first factor is the construction area of commercial buildings
  • the second factor is steel sales
  • the current time series data of the second factor includes steel sales data in January 2021
  • the second factor has three leading periods for the first factor. moon. Then, according to the steel sales data in January 2021, the prediction model based on the neural network can predict the construction area of commercial housing in 3 months.
  • the real estate index factor conforming to the real estate industry elements is closely related to multiple external environmental factors that affect the development of the real estate industry.
  • each second factor there are 500 second factors, and it is determined whether the lead period of each second factor to the first factor is greater than or equal to the preset lead period; the second factors greater than or equal to the preset lead period are reserved, and the remaining 100 are screened. about. Input the reserved second factor into the prediction model constructed for the first factor, the value of each second factor includes the lead period of each second factor to the first factor and the current time series data of each second factor , the prediction of the real estate index can be realized, and the final prediction effect is significantly better than the current real estate index prediction model.
  • the method for analyzing the real estate index data obtains the first time series data of the first factor and the second time series data of the second factor, where the first factor is the real estate index factor conforming to the real estate industry element;
  • the second time series data is seasonally adjusted and stabilized to obtain the first target time series data and the second target time series data;
  • the transfer entropy between the first target time series data and the second target time series data is calculated, and determined based on the transfer entropy
  • the first leading indicator information between the first factor and the second factor determine the linear relationship between the first target time series data and the second target time series data, and determine the first factor between the first factor and the second factor based on the linear relationship.
  • Second leading indicator information according to the first leading indicator information and the second leading indicator information, determine the target leading indicator information between the first factor and the second factor; obtain the current time series data of the second factor, and based on the target leading indicator information and Current time series data, predict the target time series data of the first factor.
  • the relevant prediction of the real estate index can be more accurately achieved, which is conducive to the accuracy of the real estate industry's expectations and helps to accurately and effectively implement the real estate industry-related regulation.
  • FIG. 4 is a schematic block diagram of an apparatus for analyzing real estate index data according to an embodiment of the present application.
  • the real estate index data analysis device 200 includes:
  • the obtaining module 201 is configured to obtain first time series data of a first factor and second time series data of a second factor, where the first factor is a real estate index factor conforming to the real estate industry element, and the second factor is an influence on the development of the real estate industry external environmental factors;
  • a processing module 202 configured to perform seasonal adjustment and smoothing processing on both the first time series data and the second time series data to obtain first target time series data and second target time series data;
  • a first determination module 203 configured to calculate the transfer entropy between the first target time series data and the second target time series data, and determine the first lead between the first factor and the second factor based on the transfer entropy indicator information;
  • a second determining module 204 configured to determine a linear relationship between the first target time series data and the second target time series data, and determine a second lead between the first factor and the second factor based on the linear relationship indicator information;
  • a third determining module 205 configured to determine target leading indicator information between the first factor and the second factor according to the first leading indicator information and the second leading indicator information;
  • the prediction module 206 is configured to acquire the current time series data of the second factor, and predict the target time series data of the first factor based on the target leading indicator information and the current time series data.
  • the first determining module 203 includes:
  • a first calculation submodule 2031 configured to calculate the amount of change of the uncertainty magnitude of the first target time series data with respect to the second target time series data, to obtain the first transfer entropy
  • the second calculation sub-module 2032 is configured to calculate the amount of change of the uncertainty size of the second target time series data with respect to the first target time series data to obtain the second transfer entropy;
  • the third calculation submodule 2033 is configured to calculate the target transfer entropy between the first target time series data and the second target time series data according to the first transfer entropy and the second transfer entropy;
  • a determination sub-module 2033 configured to determine a first lead period of the second factor for the first factor based on the target transfer entropy.
  • the processing module 202 includes:
  • the first processing sub-module 2021 is configured to process the first time series data and the second time series data respectively by using a preset ARIMA model to obtain the adjusted first time series data and the second time series data;
  • the second processing sub-module 2022 is configured to process the adjusted first time series data and the second time series data respectively through a preset empirical modal decomposition EMD model to obtain the first target time series data and the second target time series data.
  • processing module 202 is further configured to:
  • the filtered first time series data is used as the updated first target time series data
  • the filtered second target time series data is used as the updated second target time series data
  • the apparatuses provided by the above embodiments may be implemented in the form of a computer program, and the computer program may be executed on the computer device as shown in FIG. 7 .
  • FIG. 7 is a schematic block diagram of the structure of a computer device according to an embodiment of the present application.
  • the computer device can be a server or a terminal device.
  • the computer device includes a processor, a memory, and a network interface connected through a system bus, wherein the memory may include a storage medium and an internal memory.
  • the storage medium may store an operating system and a computer program.
  • the storage medium may be volatile or non-volatile.
  • the computer program includes program instructions that, when executed, can cause the processor to perform any method of analyzing real estate index data.
  • the processor is used to provide computing and control capabilities to support the operation of the entire computer equipment.
  • the internal memory provides an environment for running the computer program in the storage medium, and when the computer program is executed by the processor, the processor can cause the processor to execute any real estate index data analysis method.
  • the network interface is used for network communication, such as sending assigned tasks.
  • the network interface is used for network communication, such as sending assigned tasks.
  • FIG. 7 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer equipment to which the solution of the present application is applied. Include more or fewer components than shown in the figures, or combine certain components, or have a different arrangement of components.
  • the processor may be a central processing unit (Central Processing Unit, CPU), and the processor may also be other general-purpose processors, digital signal processors (Digital Signal Processors, DSP), application specific integrated circuits (Application Specific Integrated circuits) Circuit, ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.
  • the general-purpose processor can be a microprocessor or the processor can also be any conventional processor or the like.
  • the processor is configured to run a computer program stored in the memory to implement the following steps:
  • first time series data of a first factor and second time series data of a second factor, where the first factor is a real estate index factor conforming to the real estate industry element, and the second factor is an external environmental factor affecting the development of the real estate industry;
  • the processor performs the calculation of the transfer entropy between the first target time series data and the second target time series data, and determines the first factor and the second factor based on the transfer entropy
  • the first leading indicator information between is used to achieve:
  • the first transfer entropy and the second transfer entropy calculate the target transfer entropy between the first target time series data and the second target time series data
  • a first lead term of the second factor with respect to the first factor is determined.
  • the processor performs the determining of a linear relationship between the first target time series data and the second target time series data, and determines the first factor and the second factor based on the linear relationship
  • the second leading indicator information between is used to achieve:
  • a plurality of lag term coefficients between the first target time series data and the second target time series data are calculated, and the second factor is determined based on the plurality of lag term coefficients for the The second lead period for the first factor.
  • the processor performs the seasonal adjustment and smoothing processing on both the first time series data and the second time series data to obtain the first target time series data and the second target time series data, Also used to implement:
  • the step of calculating the transfer entropy between the first target time series data and the second target time series data is performed;
  • the target leading indicator information between the first factor and the second factor is determined through a preset Barzilai-Borwein gradient algorithm.
  • the processor when the processor implements the predetermined Barzilai-Borwein gradient algorithm to determine the target leading indicator information between the first factor and the second factor, the processor is configured to implement:
  • the processor implements the seasonal adjustment and smoothing processing on both the first time series data and the second time series data to obtain the first target time series data and the second target time series data, Used to implement:
  • the first time series data and the second time series data are respectively processed by the preset ARIMA model to obtain the adjusted first time series data and the second time series data;
  • the adjusted first time series data and the second time series data are processed respectively through a preset empirical modal decomposition EMD model to obtain first target time series data and second target time series data.
  • the processor performs the seasonal adjustment and smoothing processing on both the first time series data and the second time series data to obtain the first target time series data and the second target time series data, Also used to implement:
  • the first target time series data and the second target time series data are filtered respectively to obtain the filtered first time series data and the filtered second target time series;
  • the filtered first time series data is used as the updated first target time series data
  • the filtered second target time series data is used as the updated second target time series data
  • the first time series data of the first factor and the second time series data of the second factor are obtained, where the first factor is a real estate index factor conforming to the real estate industry element; seasonality is performed on both the first time series data and the second time series data Adjustment and smoothing process to obtain the first target time series data and the second target time series data; calculate the transfer entropy between the first target time series data and the second target time series data, and determine the difference between the first factor and the second factor based on the transfer entropy.
  • the first leading indicator information between the two determines the linear relationship between the first target time series data and the second target time series data, and determine the second leading indicator information between the first factor and the second factor based on the linear relationship;
  • Leading indicator information and second leading indicator information determine the target leading indicator information between the first factor and the second factor; obtain the current time series data of the second factor, and predict the first factor based on the target leading indicator information and the current time series data target time series data.
  • Embodiments of the present application further provide a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, wherein when the computer program is executed by a processor, the following steps are implemented:
  • first time series data of a first factor and second time series data of a second factor, where the first factor is a real estate index factor conforming to the real estate industry element, and the second factor is an external environmental factor affecting the development of the real estate industry;
  • the computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as a hard disk or a memory of the computer device.
  • the computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk equipped on the computer device, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) ) card, Flash Card, etc.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • General Physics & Mathematics (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Educational Administration (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • General Health & Medical Sciences (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Fuzzy Systems (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种房产指数数据分析方法、装置、设备及存储介质,该方法包括:通过获取涉及房产指数的第一因子的第一时序数据和第二因子的第二时序数据;对第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;计算第一目标时序数据和第二目标时序数据之间的传递熵和线性关系,并基于该传递熵和线性关系,确定第一因子与第二因子之间的目标领先指标信息;获取第二因子的当前时序数据,并基于目标领先指标信息和当前时序数据,预测第一因子的目标时序数据。该方法能够提高房产指数分析预测的准确性。

Description

房产指数数据分析方法、装置、设备及存储介质
本申请要求于2021年4月30日提交中国专利局、申请号为202110486524.6、发明名称为“房产指数数据分析方法、装置、设备及存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及数据分析的技术领域,尤其涉及一种房产指数数据分析方法、装置、设备及存储介质。
背景技术
房地产业是指从事土地和房地产开发、经营、管理和服务的行业。目前,影响房地产业发展的外部环境因子众多,例如消费者物价指数(Consumer Price Index,CPI)、生产价格指数(Producer Price Index,PPI)、钢铁销量、家电出口、住房公积金贷款率、开工面积和去化周期等,这些外部环境因子与房地产的销售价格、销售面积等符合房地产业要素的房产指数密切相关。发明人意识到当前房地产行业大多采用统计方法进行外部环境因子与房产指数之间的相关性预测,比如通过循环设定滞后阶数计算外部环境因子对房产指数的相关预测,通过这种方式预测的房产指数并不符合房地产行业和宏观经济周期的规律,导致外部环境因子对房产指数的相关预测并不准确,不利于政府准确有效地实施房地产调控,难以保证后续房地产行业预期的准确性。
发明内容
本申请的主要目的在于提供一种房产指数数据分析方法、装置、设备及存储介质,旨在提高房产指数分析预测的准确性,从而有利于准确且有效地实施房地产相关调控。
第一方面,本申请提供一种房产指数数据分析方法,包括:
获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
第二方面,本申请还提供一种房产指数数据分析装置,所述房产指数数据分析装置包括:
获取模块,用于获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
处理模块,用于对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
第一确定模块,用于计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
第二确定模块,用于确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
第三确定模块,用于根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
预测模块,用于获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
第三方面,本申请还提供一种计算机设备,所述计算机设备包括处理器、存储器、以及存储在所述存储器上并可被所述处理器执行的计算机程序,其中所述计算机程序被所述处理器执行时,实现以下步骤:
获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
第四方面,本申请还提供一种计算机可读存储介质,所述计算机可读存储介质上存储有计算机程序,其中所述计算机程序被处理器执行时,实现以下步骤:
获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
本申请提供一种房产指数数据分析方法、装置、设备及存储介质,本申请通过获取第一因子的第一时序数据和第二因子的第二时序数据,第一因子为符合房地产业要素的房产指数因子;对第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;计算第一目标时序数据和第二目标时序数据之间的传递熵,并基于传递熵确定第一因子与第二因子之间的第一领先指标信息;确定第一目标时序数据和第二目标时序数据之间的线性关系,并基于线性关系确定第一因子与第二因子之间的第二领先指标信息;根据第一领先指标信息和第二领先指标信息,确定第一因子与第二因子之间的目标领先指标信息;获取第二因子的当前时序数据,并基于目标领先指标信息 和当前时序数据,预测第一因子的目标时序数据。通过目标领先指标信息能够较为准确地实现房产指数的相关分析预测,有利于房地产行业预期的准确性,有助于准确有效地实施房地产业相关调控。
附图说明
为了更清楚地说明本申请实施例技术方案,下面将对实施例描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1为本申请实施例提供的一种房产指数数据分析方法的步骤流程示意图;
图2为图1中的房产指数数据分析方法的一子步骤流程示意图;
图3为图1中的房产指数数据分析方法的另一子步骤流程示意图;
图4为本申请实施例提供的一种房产指数数据分析装置的示意性框图;
图5为图4中的房产指数数据分析装置的一子模块的示意性框图;
图6为图4中的房产指数数据分析装置的另一子模块的示意性框图;
图7为本申请实施例提供的一种计算机设备的结构示意性框图。
本申请目的的实现、功能特点及优点将结合实施例,参照附图做进一步说明。
具体实施方式
下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
本申请实施例提供一种房产指数数据分析方法、装置、设备及存储介质。其中,该房产指数数据分析方法可应用于终端设备或服务器中,该终端设备可以为手机、平板电脑、笔记本电脑、台式电脑、个人数字助理和穿戴式设备等电子设备;该服务器可以为单台的服务器,也可以为由多台服务器组成的服务器集群。以下以该房产指数数据分析方法应用于服务器为例进行解释说明。
下面结合附图,对本申请的一些实施方式作详细说明。在不冲突的情况下,下述的实施例及实施例中的特征可以相互组合。
请参照图1,图1为本申请实施例提供的一种房产指数数据分析方法的步骤流程示意图。
如图1所示,该房产指数数据分析方法包括步骤S101至步骤S106。
步骤S101、获取第一因子的第一时序数据和第二因子的第二时序数据。
其中,第一因子为符合房地产业要素的房产指数因子,第一因子可以是频率比较低的具有经济周期波动和宏观意义上与房地产行业相关的因子,例如第一因子为各地区商品房的销售价格、销售面积或者销售量增长率,又例如第一因子为各地区二手房交易价格、交易量或者交易增长情况等。第二因子为影响房地产业发展的外部环境因子,例如第二因子为消费者物价指数(Consumer Price Index,CPI)、生产价格指数(Producer Price Index,PPI)、钢铁销量、家电出口、住房公积金贷款率、开工面积或者去化周期等。
其中,第一因子的第一时序数据可以是同一房产指数因子统一按时间顺序记录的数据列,在同一数据列中的各个数据是同口径的,具有可比性。第一时序数据可以是时期数,也可以时点数。同理,第二因子的第二时序数据可以是同一外部环境因子统一按时间顺序记录的数据列,可以是时期数,也可以时点数,本实施例不做具体限定。
在一实施例中,服务器从存储器或者云端中获取第一因子的第一时序数据以及第二因子的第二时序数据,或者分别从不同的存储器或者数据库中获取第一因子的第一时序数据以及第二因子的第二时序数据,例如从第一数据库中获取第一因子的第一时序数据,并从 第二数据库中获取第二因子的第二时序数据,本实施例对此不做具体限定。
需要说明的是,为进一步保证上述第一时序数据和第二时序数据等相关信息的私密和安全性,上述第一时序数据和第二时序数据等相关信息还可以存储于一区块链的节点中,本申请的技术方案还可适用于添加其他存储于区块链上的数据文件,本申请所指区块链是分布式数据存储、点对点传输、共识机制、加密算法等计算机技术的新型应用模式。区块链(Blockchain),本质上是一个去中心化的数据库,是一串使用密码学方法相关联产生的数据块,每一个数据块中包含了一批次网络交易的信息,用于验证其信息的有效性(防伪)和生成下一个区块。区块链可以包括区块链底层平台、平台产品服务层以及应用服务层等。
步骤S102、对第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据。
需要说明的是,相较于以往的房产指数因子的时序数据处理,本实施例对时序数据的处理维度更加全面。通过对第一时序数据和第二时序数据进行季节性调整,纠正第一时序数据和第二时序数据中隐含季节因素的影响,让不同季节的第一时序数据和第二时序数据具有可比性,令经过调整的第一目标时序数据和第二目标时序数据更能反映出实际的变化趋势。通过对第一时序数据和第二时序数据进行平稳化处理,能够更好的提取第一时序数据和第二时序数据中的经济意义,令得到的第一目标时序数据和第二目标时序数据更加平稳,进一步保证数据有效性以及后续房产指数分析预测的准确性。
在一实施例中,如图2所示,步骤S102包括:子步骤S1021至子步骤S1022。
子步骤S1021、通过预设的ARIMA模型分别对第一时序数据和第二时序数据进行处理,得到经过调整的第一时序数据和第二时序数据。
其中,ARIMA模型包括考虑日历、节假日和交易日等的reg ARIMA模型或季节ARIMA模型。通过ARIMA模型分别对第一时序数据和第二时序数据进行预处理,纠正第一时序数据和第二时序数据中隐含季节因素的影响,让不同季节的时序数据具有可比性,令经过调整的第一时序数据和第二时序数据更能反映出实际的变化趋势,从而提高房产指数分析预测的准确性。
子步骤S1022、通过预设的经验模态分解EMD模型分别对经过调整的第一时序数据和第二时序数据进行处理,得到第一目标时序数据和第二目标时序数据。
其中,经验模态分解EMD模型依据经过调整的第一时序数据和第二时序数据自身的时间尺度特征来进行信号分解,无须预先设定任何基函数。示例性的,通过令经过调整的第一时序数据和第二时序数据各自减去极大值与极小值,并分别进行三次样条插值后得到的包络线平均值令第一目标时序数据和第二目标时序数据平稳,从而能够更好的提取第一时序数据和第二时序数据中的经济意义,进一步保证第一目标时序数据和第二目标时序数据的有效性,从而提高房产指数分析预测的准确性。
进一步地,对第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据之后,还包括:分别对第一目标时序数据和第二目标时序数据进行滤波处理,得到过滤的第一时序数据和过滤的第二目标时序;将过滤的第一时序数据作为更新的第一目标时序数据,并将过滤的第二目标时序作为更新的第二目标时序数据。需要说明的是,通过对第一目标时序数据和第二目标时序数据分别进行滤波处理,能够防止第一目标时序数据和第二目标时序数据中的噪声对于后续建模的产生消极的影响,通过本实施例采用的滤波方法分别对于第一目标时序数据和第二目标时序数据中的噪声进行过滤,剔除无效异常数据的影响,进一步保证数据的有效性以及保证房产指数分析预测的准确性。
在一实施例中,对第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据之后,还包括:对第二目标时序数据进行平稳性检验,得到检验结果;若检验结果为第二目标时序数据平稳,则执行计算第一目标时序数 据和第二目标时序数据之间的传递熵的步骤;若检验结果为第二目标时序数据不平稳,则通过预设的Barzilai-Borwein梯度算法,确定第一因子与第二因子之间的目标领先指标信息。其中,可采用单位根ADF检验方法对目标时序数据进行平稳性检验,例如假设第二目标时序数据平稳,并计算p值,若p值小于0.05,则拒绝原假设。需要说明的是,经过季节性调整和平稳化处理后,第二目标时序数据不一定平稳,通过对第二目标时序数据进行平稳性检验,并根据检验结果选取不同的实施策略来计算第一因子与第二因子之间的目标领先指标信息,目标领先指标信息更加准确,有利于进一步提高房产指数分析预测的准确性。
通过实验证明,如果第二目标时序序列不平稳,直接使用Barzilai-Borwein(BB)梯度算法确定第一因子与第二因子之间的目标领先指标信息,得到的目标领先指标信息更加准确,但如果第二目标时序数据平稳,可执行步骤S103及其之后的步骤来计算第一因子与第二因子之间的目标领先指标信息,通过离散化计算使得到的目标领先指标信息更加准确。
进一步地,通过预设的Barzilai-Borwein梯度算法,确定第一因子与第二因子之间的第二领先指标信息,包括:分别对第一目标时序数据和第二目标时序数据进行分段线性处理,以确定第一目标时序数据的第一波峰和第一波谷,以及确定第二目标时序数据的第二波峰和第二波谷;从多个第二波峰中确定在第一波谷之前的目标波峰,并计算目标波峰与第一波谷之间的时间差值;并从多个第二波谷中确定在第一波峰之前的目标波谷,并计算目标波谷与第一波峰之间的时间差值;根据目标波峰与第一波谷之间的时间差值、以及目标波谷与第一波峰之间的时间差值,确定第二因子对于第一因子的第一领先期限。
需要说明的是,通过对第一目标时序数据和第二目标时序数据进行分段线性处理,能够得到第一目标时序数据和第二目标时序数据向上或者向下的趋势。以第一目标时序数据为例,对第一目标时序数据进行分段线性处理后,得到多个极值点(包括极大值点和极小值点),从第一个极值点开始迭代,要求相邻两个极值点方向相反,即如果第一个是波峰,下一个线性化的端点一定要是波谷,如果不是则认为当前端点为拐点,不是真正的波峰或波谷,从而确定第一目标时序数据的第一波峰和第一波谷。相类似的,第二目标时序数据的第二波峰和第二波谷也可通过分段线性处理得到。
需要说明的是,将第一波谷和第二波峰进行匹配,从多个第二波峰中确定在第一波谷之前的目标波峰,即选取在第一波谷之前的多个第二波峰中的目标波峰。其中,也可从多个第二波峰中确定在第一波谷预设时间范围之前的目标波峰,预设时间范围可灵活设置,例如预设时间范围为6个月,即选取在第一波谷之前6个月内的多个第二波峰中的目标波峰。相类似的,可以用同样的方式从多个第二波谷中确定在第一波峰之前的目标波谷。
需要说明的是,对第二目标时序数据而言,用匹配的目标波峰的数量除以第二波谷的总数量得到的值,认为是与第一目标时序数据之间的匹配概率,根据该匹配概率和预设阈值确定第二因子是否对第一因子具有领先作用。示例性的,预设阈值为0.7,如果匹配概率低于0.7则认为两个序列不匹配,即第二因子不具有领先作用;如果匹配概率大于等于0.7认为第一目标时序数据与第二目标时序数据是匹配的,此时认为第二因子对第一因子具有领先作用。通过计算目标波峰与第一波谷之间的时间差值、以及目标波谷与第一波峰之间的时间差值的平均值,确定第二因子对于第一因子的第一领先期限。
在一实施例中,若第二因子对第一因子具有领先作用,则确定第二因子对于第一因子的第一领先期限。第一领先期限为第一目标时序数据与第二目标时序数据中相互匹配的波峰和波谷时间之差的平均值。具体地,根据目标波峰与第一波谷之间的时间差值、以及目标波谷与第一波峰之间的时间差值,确定第二因子对于第一因子的第一领先期限,包括:计算目标波峰与每个第一波谷之间的时间差值,得到多个第一时间差值;计算目标波谷与每个第一波峰之间的时间差值,得到多个第二时间差值;计算多个第一时间差值与多个第二时间差值的平均值,得到第二因子对于第一因子的第一领先期限。
步骤S103、计算第一目标时序数据和第二目标时序数据之间的传递熵,并基于传递熵确定第一因子与第二因子之间的第一领先指标信息。
其中,信息熵用于衡量变量不确定性的大小,考虑的是变量间的信息量传递,而不需要假定变量间具有特定形式的关系,尤其是对于非线性的时间序列数据。第一领先指标信息包括第一因子与第二因子之间的领先滞后关系,例如第二因子对于第一因子的领先期限或者滞后期限。需要说明的是,基于传递熵确定第一因子与第二因子之间的第一领先指标信息,便于捕捉第一目标时序数据和第二目标时序数据之间的非线性的领先关系。把单向的传递熵的概念用在了时间序列数据上,有利于构建净传递比消除量纲的影响,从而提高房产指数分析预测的准确性。
在一实施例中,如图3所示,步骤S103包括:子步骤S1031至子步骤S1034。
子步骤S1031、计算第一目标时序数据对于第二目标时序数据的不确定性大小的改变量,得到第一传递熵。
其中,传递熵的计算公式可以为:
Figure PCTCN2021097416-appb-000001
p、q分别表示两个数据分布。
根据现有定义,传递熵表示y的信息对于x的信息的不确定性的改变,本实施例创新地将传递熵的概念应用于时间序列上,捕捉两个目标时序数据之间的非线性的领先关系。因此,若将第一目标时序数据记为Xn,并将第二目标时序数据记为Yn,则Xn-t时刻对Yn有第一传递熵记为H(x->y),其中,t为时间度量。
子步骤S1032、计算第二目标时序数据对于第一目标时序数据的不确定性大小的改变量,得到第二传递熵。
需要说明的是,为了更好地确定第二目标时序数据与第一目标时序数据之间的领先滞后关系,还需要确定第二目标时序数据对于第一目标时序数据的不确定性大小的改变量,便于排除单向传递熵本身的量纲影响。
因此,若将第一目标时序数据记为Xn,并将第二目标时序数据记为Yn,同时计算了Yn-t时刻到Xn的信息传递,第二传递熵记为H(y->x)。
子步骤S1033、根据第一传递熵和第二传递熵,计算第一目标时序数据和第二目标时序数据之间的目标传递熵。
需要说明的是,通过实验发现存在第一目标时序数据Xn和第二目标时序数据Yn相互影响的情况,为了排除掉不是真正存在领先的第二因子,消除单向传递熵的量纲影响,需要根据第一传递熵和第二传递熵,计算目标传递熵。
在一实施例中,根据第一传递熵和第二传递熵,确定第二因子与第一因子之间的领先滞后关系。计算第一传递熵与第二传递熵之间的差值H(x->y)-H(y->x),若该差值为正值,则说明第二因子对于第一因子存在领先关系,若该差值为负值,则说明第二因子对于第一因子存在滞后关系。
在一实施例中,目标传递熵的计算公式为:
Figure PCTCN2021097416-appb-000002
其中,目标传递熵为TE,第一传递熵为H1,第二传递熵为H2。φ为了防止分母为零,可取0.00001,abs(H1)为第一传递熵H1的绝对值,abs(H2)为第二传递熵H2的绝对值。通过计算第一目标时序数据和第二目标时序数据之间的目标传递熵,能够消除单向传递熵本身的量纲影响,便于准确地确定第二因子对于第一因子的第一领先期限,从而提高房产指数分析预测的准确性,从而有利于准确且有效地实施房地产相关调控。
子步骤S1034、基于目标传递熵,确定第二因子对于第一因子的第一领先期限。
在一实施例中,确定目标传递熵是否大于预设净传递比值,目标传递熵的值域为(0,1),根据该目标传递熵和预设净传递比值确定第二因子是否对第一因子具有领先作用。 可选的,预设净传递比值为0.5。如果目标传递熵低于预设净传递比值,则第二因子不具有领先作用;如果目标传递熵大于等于预设净传递比值,则第二因子对第一因子具有领先作用。
在一实施例中,第一传递熵H(x->y)包括Xn-t时刻对Yn的信息传递,则第一传递熵有多个,第二传递熵H(y->x)包括Yn-t时刻到Xn的信息传递,则第二传递熵有多个,其中,t为时间度量,例如日、周、月或者年。同理可得,目标传递熵有多个。选取最大的目标传递熵对应的第一目标传递熵或第二目标传递熵,并确定第一目标传递熵或第二目标传递熵对应的时间序列差值,得到第二因子对于第一因子的第一领先期限,其中,该时间序列差值为第一因子的第一时序数据以及第二因子的第二时序数据之间的时间差值。
示例性的,时间度量t为月,第一时序数据包括2020年1月至12月的房产指数因子的时间序列,第二时序数据包括2020年1月至12月的外部环境因子的时间序列。最大的目标传递熵为0.8,其对应的第一目标传递熵是使用2020年6月的第一时序数据与2020年10月的第二时序数据计算得到的,即第一目标传递熵对应的时间序列差值为2020年10月与2020年6月的时间差值,得到外部环境因子对于房产指数因子的时间差值为4个月,即第二因子对于第一因子的第一领先期限为4个月。
步骤S104、确定第一目标时序数据和第二目标时序数据之间的线性关系,并基于线性关系确定第一因子与第二因子之间的第二领先指标信息。
其中,第二领先指标信息包括第一因子与第二因子之间的领先滞后关系,例如第二因子对于第一因子的领先期限或者滞后期限。需要说明的是,由于前述计算第一目标时序数据和第二目标时序数据之间的传递熵对数据进行了离散,对线性关系的捕捉稍弱于线性相关性,第一领先指标信息的输出结果偏向于包含较多的非线性相关因子。因此,确定第一目标时序数据和第二目标时序数据之间的线性关系,并基于线性关系确定第一因子与第二因子之间的第二领先指标信息,有利于全面整合第一目标时序数据和第二目标时序数据之间的线性关系和非线性关系,提高房产指数分析和预测的准确性。
在一实施例中,基于预设的格兰杰因果检验算法,计算第一目标时序数据和第二目标时序数据之间的因果关系,并基于因果关系确定第二因子对于第一因子的第二领先期。需要说明的是,通过格兰杰因果检验算法研究第一目标时序数据和第二目标时序数据的历史值对于预测值的准确性,得到第一目标时序数据和第二目标时序数据之间的统计学因果关系,并基于统计学因果关系确定第二因子对于第一因子的第二领先期,补充了第一目标时序数据和第二目标时序数据之间的非线性的相关关系,能够提高房产指数分析和预测的准确性。
其中,格兰杰因果检验的回归公式为:
Figure PCTCN2021097416-appb-000003
其中,u 1t为白噪音,α、β为设定系数,y为第一因子的第一目标时序数据,x为第二因子的第二目标时序数据。
在一实施例中,基于预设的滞后相关系数算法,计算第一目标时序数据和第二目标时序数据之间的多个滞后项系数,并基于多个滞后项系数,确定第二因子对于第一因子的第二领先期。需要说明的是,可从多个滞后项系数中选取最大的滞后项系数作为第二因子对于第一因子的领先期,或者计算多个滞后项系数的平均值,并将多个滞后项系数的平均值作为领先期。
其中,滞后相关系数算法的计算公式为:
Figure PCTCN2021097416-appb-000004
其中,r l为滞后项系数,y为第一因子的第一目标时序数据,x为第二因子的第二目标时序数据。
Figure PCTCN2021097416-appb-000005
为第一目标时序数据的平均值,
Figure PCTCN2021097416-appb-000006
为第二目标时序数据的平均值。
在一实施例中,基于预设的格兰杰因果检验算法和滞后相关系数算法对第一目标时序数据和第二目标时序数据进行处理,以确定第二因子对于第一因子的第二领先期。可以理解的是,也可以采用本领域内的其他线性相关的算法或者模型来计算第一因子与第二因子之间的第二领先指标信息,本实施例不做具体限定。
步骤S105、根据第一领先指标信息和第二领先指标信息,确定第一因子与第二因子之间的目标领先指标信息。
其中,目标领先指标信息包括第一因子与第二因子之间的领先滞后关系,例如第二因子对于第一因子的领先期限或者滞后期限。需要说明的是,得到第一领先指标信息的处理过程进行了离散,第一领先指标信息偏向于包含较多的非线性相关因子,对非线性关系的捕捉较强。得到第二领先指标信息的处理过程强调对线性相关因子的捕捉,倾向于筛选出很多线性相关的因子,本申请最终的目标是筛选出同时具有线性关系和非线性关系这两种性质的因子,所以可根据第一领先指标信息和第二领先指标信息,确定目标领先指标信息,能够极大地提高房产指数分析和预测的准确性。
在一实施例中,对第一领先指标信息和第二领先指标信息取交集,得到第一因子与第二因子之间的目标领先指标信息。例如,第一领先指标信息为第二因子对于第一因子的第一领先期限为4个月,第二领先指标信息为第二因子对于第一因子的第二领先期限为2个月,对该第一领先指标信息和第二领先指标信息取交集,则目标领先指标信息为第二因子对于第一因子的领先期限为2个月至4个月。
在一实施例中,第一领先指标信息和第二领先指标信息均包括第二因子与第一因子之间的领先滞后关系的时间范围,可根据第一领先指标信息和第二领先指标信息,确定第二因子对于第一因子的领先期限。例如,第一领先指标信息为第二因子对于第一因子的第一领先期限为1-4个月,第二领先指标信息为第二因子对于第一因子的第二领先期限为0-2个月,对该第一领先指标信息和第二领先指标信息取交集,则目标领先指标信息为第二因子对于第一因子的领先期限为1-2个月。本实施例不做具体限定。
实践证明,第一因子与第二因子之间的目标领先指标信息的计算方法越多,该计算方法包括计算传递熵算法、交叉熵算法等线性相关算法、以及格兰杰因果检验算法、滞后相关系数算法等非线性相关算法,第一领先指标信息和第二领先指标信息的公共交集越小,噪声的影响被扩大,目标领先指标期的范围越小。
步骤S106、获取第二因子的当前时序数据,并基于目标领先指标信息和当前时序数据,预测第一因子的目标时序数据。
需要说明的是,目标领先指标信息包括第二因子对于第一因子的领先期限,基于第二因子对于第一因子的领先期限和第二因子的当前时序数据,可以预测在第二因子对于第一因子的领先期限之后的第一因子的目标时序数据。从而能够基于该目标时序数据针对性地实施房地产调控,保证房地产行业能够按照预期发展。
示例性的,第一因子为商品房的建筑面积,第二因子为钢铁销量,第二因子的当前时序数据包括2021年1月份的钢铁销量数据,第二因子对于第一因子的领先期限为3个月。那么,根据2021年1月份的钢铁销量数据,基于神经网络组成的预测模型,能够预测3个月后的商品房的建筑面积。
在一实施例中,第二因子为多个,符合房地产业要素的房产指数因子是由多个影响房地产业发展的外部环境因子密切相关。从多个第二因子中筛选出对第一因子具有领先作用的多个目标因子;获取每个目标因子的当前时序数据,并基于每个目标因子的目标领先指标信息和每个目标因子的当前时序数据,共同预测第一因子的目标时序数据。需要说明的是,通过对第一因子具有领先作用的多个目标因子,共同预测第一因子的目标时序数据, 能够极大提高房产指数分析预测的准确性,从而有利于准确且有效地实施房地产相关调控。
示例性的,第二因子有500个,确定每个第二因子对第一因子的领先期限是否大于或等于预设领先期限;保留大于或等于预设领先期限的第二因子,筛选余下100个左右。将保留的该第二因子输入至针对第一因子而构建的预测模型,每个第二因子的取值包括每个第二因子对第一因子的领先期限和每个第二因子的当前时序数据,即可实现对房产指数的预测,最终的预测效果显著优于当前的房产指数预测模型。
上述实施例提供的房产指数数据分析方法,通过获取第一因子的第一时序数据和第二因子的第二时序数据,第一因子为符合房地产业要素的房产指数因子;对第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;计算第一目标时序数据和第二目标时序数据之间的传递熵,并基于传递熵确定第一因子与第二因子之间的第一领先指标信息;确定第一目标时序数据和第二目标时序数据之间的线性关系,并基于线性关系确定第一因子与第二因子之间的第二领先指标信息;根据第一领先指标信息和第二领先指标信息,确定第一因子与第二因子之间的目标领先指标信息;获取第二因子的当前时序数据,并基于目标领先指标信息和当前时序数据,预测第一因子的目标时序数据。通过目标领先指标信息能够较为准确地实现房产指数的相关预测,有利于房地产行业预期的准确性,有助于准确有效地实施房地产业相关调控。
请参照图4,图4为本申请实施例提供的一种房产指数数据分析装置的示意性框图。
如图4所示,该房产指数数据分析装置200,包括:
获取模块201,用于获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
处理模块202,用于对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
第一确定模块203,用于计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
第二确定模块204,用于确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
第三确定模块205,用于根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
预测模块206,用于获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
在一个实施例中,如图5所示,第一确定模块203包括:
第一计算子模块2031,用于计算所述第一目标时序数据对于所述第二目标时序数据的不确定性大小的改变量,得到第一传递熵;
第二计算子模块2032,用于计算所述第二目标时序数据对于所述第一目标时序数据的不确定性大小的改变量,得到第二传递熵;
第三计算子模块2033,用于根据所述第一传递熵和所述第二传递熵,计算所述第一目标时序数据和第二目标时序数据之间的目标传递熵;
确定子模块2033,用于基于所述目标传递熵,确定所述第二因子对于所述第一因子的第一领先期限。
在一个实施例中,如图6所示,处理模块202包括:
第一处理子模块2021,用于通过预设的ARIMA模型分别对所述第一时序数据和第二时序数据进行处理,得到经过调整的所述第一时序数据和所述第二时序数据;
第二处理子模块2022,用于通过预设的经验模态分解EMD模型分别对经过调整的所述第一时序数据和所述第二时序数据进行处理,得到第一目标时序数据和第二目标时序数 据。
在一个实施例中,处理模块202还用于:
分别对所述第一目标时序数据和第二目标时序数据进行滤波处理,得到过滤的第一时序数据和过滤的第二目标时序;
将所述过滤的第一时序数据作为更新的第一目标时序数据,并将所述过滤的第二目标时序作为更新的第二目标时序数据。
需要说明的是,所属领域的技术人员可以清楚地了解到,为了描述的方便和简洁,上述描述的装置和各模块及单元的具体工作过程,可以参考前述房产指数数据分析方法实施例中的对应过程,在此不再赘述。
上述实施例提供的装置可以实现为一种计算机程序的形式,该计算机程序可以在如图7所示的计算机设备上运行。
请参阅图7,图7为本申请实施例提供的一种计算机设备的结构示意性框图。该计算机设备可以为服务器或终端设备。
如图7所示,该计算机设备包括通过系统总线连接的处理器、存储器和网络接口,其中,存储器可以包括存储介质和内存储器。
存储介质可存储操作系统和计算机程序。所述存储介质可以是易失性的,也可以是非易失性的。该计算机程序包括程序指令,该程序指令被执行时,可使得处理器执行任意一种房产指数数据分析方法。
处理器用于提供计算和控制能力,支撑整个计算机设备的运行。
内存储器为存储介质中的计算机程序的运行提供环境,该计算机程序被处理器执行时,可使得处理器执行任意一种房产指数数据分析方法。
该网络接口用于进行网络通信,如发送分配的任务等。本领域技术人员可以理解,图7中示出的结构,仅仅是与本申请方案相关的部分结构的框图,并不构成对本申请方案所应用于其上的计算机设备的限定,具体的计算机设备可以包括比图中所示更多或更少的部件,或者组合某些部件,或者具有不同的部件布置。
应当理解的是,处理器可以是中央处理单元(Central Processing Unit,CPU),该处理器还可以是其他通用处理器、数字信号处理器(Digital Signal Processor,DSP)、专用集成电路(Application Specific Integrated Circuit,ASIC)、现场可编程门阵列(Field-Programmable Gate Array,FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。其中,通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
其中,在一个实施例中,所述处理器用于运行存储在存储器中的计算机程序,以实现如下步骤:
获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
在一个实施例中,所述处理器在实现所述计算所述第一目标时序数据和第二目标时序 数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息时,用于实现:
计算所述第一目标时序数据对于所述第二目标时序数据的不确定性大小的改变量,得到第一传递熵;
计算所述第二目标时序数据对于所述第一目标时序数据的不确定性大小的改变量,得到第二传递熵;
根据所述第一传递熵和所述第二传递熵,计算所述第一目标时序数据和第二目标时序数据之间的目标传递熵;
基于所述目标传递熵,确定所述第二因子对于所述第一因子的第一领先期限。
在一个实施例中,所述处理器在实现所述确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息时,用于实现:
基于预设的格兰杰因果检验算法,计算所述第一目标时序数据和第二目标时序数据之间的因果关系,并基于所述因果关系确定所述第二因子对于所述第一因子的第二领先期;或者
基于预设的滞后相关系数算法,计算所述第一目标时序数据和第二目标时序数据之间的多个滞后项系数,并基于所述多个滞后项系数确定所述第二因子对于所述第一因子的第二领先期。
在一个实施例中,所述处理器在实现所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据之后,还用于实现:
对所述第二目标时序数据进行平稳性检验,得到检验结果;
若检验结果为第二目标时序数据平稳,则执行所述计算所述第一目标时序数据和第二目标时序数据之间的传递熵的步骤;
若检验结果为第二目标时序数据不平稳,则通过预设的Barzilai-Borwein梯度算法,确定所述第一因子与第二因子之间的目标领先指标信息。
在一个实施例中,所述处理器在实现所述通过预设的Barzilai-Borwein梯度算法,确定所述第一因子与第二因子之间的目标领先指标信息时,用于实现:
分别对所述第一目标时序数据和第二目标时序数据进行分段线性处理,以确定所述第一目标时序数据的第一波峰和第一波谷,以及确定所述第二目标时序数据的第二波峰和第二波谷;
从多个所述第二波峰中确定在所述第一波谷之前的目标波峰,并计算所述目标波峰与所述第一波谷之间的时间差值;
并从多个所述第二波谷中确定在所述第一波峰之前的所述目标波谷,并计算所述目标波谷与所述第一波峰之间的时间差值;
根据所述目标波峰与所述第一波谷之间的时间差值、以及所述目标波谷与所述第一波峰之间的时间差值,确定所述第二因子对于所述第一因子的目标领先期限。
在一个实施例中,所述处理器在实现所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据时,用于实现:
通过预设的ARIMA模型分别对所述第一时序数据和第二时序数据进行处理,得到经过调整的所述第一时序数据和所述第二时序数据;
通过预设的经验模态分解EMD模型分别对经过调整的所述第一时序数据和所述第二时序数据进行处理,得到第一目标时序数据和第二目标时序数据。
在一个实施例中,所述处理器在实现所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据之后,还用于实现:
分别对所述第一目标时序数据和第二目标时序数据进行滤波处理,得到过滤的第一时 序数据和过滤的第二目标时序;
将所述过滤的第一时序数据作为更新的第一目标时序数据,并将所述过滤的第二目标时序作为更新的第二目标时序数据。
需要说明的是,所属领域的技术人员可以清楚地了解到,为了描述的方便和简洁,上述描述计算机设备的具体工作过程,可以参考前述房产指数数据分析方法实施例中的对应过程,在此不再赘述。
本实施例通过获取第一因子的第一时序数据和第二因子的第二时序数据,第一因子为符合房地产业要素的房产指数因子;对第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;计算第一目标时序数据和第二目标时序数据之间的传递熵,并基于传递熵确定第一因子与第二因子之间的第一领先指标信息;确定第一目标时序数据和第二目标时序数据之间的线性关系,并基于线性关系确定第一因子与第二因子之间的第二领先指标信息;根据第一领先指标信息和第二领先指标信息,确定第一因子与第二因子之间的目标领先指标信息;获取第二因子的当前时序数据,并基于目标领先指标信息和当前时序数据,预测第一因子的目标时序数据。通过目标领先指标信息能够较为准确地实现房产指数的相关预测,有利于房地产行业预期的准确性,有助于准确有效地实施房地产业相关调控。
本申请实施例还提供一种计算机可读存储介质,所述计算机可读存储介质上存储有计算机程序,其中所述计算机程序被处理器执行时,实现以下步骤:
获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
需要说明的是,所属领域的技术人员可以清楚地了解到,为了描述的方便和简洁,上述描述计算机可读存储介质的具体工作过程,可以参考前述房产指数数据分析方法实施例中的对应过程,在此不再赘述。
其中,所述计算机可读存储介质可以是前述实施例所述的计算机设备的内部存储单元,例如所述计算机设备的硬盘或内存。所述计算机可读存储介质也可以是所述计算机设备的外部存储设备,例如所述计算机设备上配备的插接式硬盘,智能存储卡(Smart Media Card,SMC),安全数字(Secure Digital,SD)卡,闪存卡(Flash Card)等。
上述本申请实施例序号仅仅为了描述,不代表实施例的优劣。以上所述,仅为本申请的具体实施方式,但本申请的保护范围并不局限于此,任何熟悉本技术领域的技术人员在本申请揭露的技术范围内,可轻易想到各种等效的修改或替换,这些修改或替换都应涵盖在本申请的保护范围之内。因此,本申请的保护范围应以权利要求的保护范围为准。

Claims (20)

  1. 一种房产指数数据分析方法,其中,包括:
    获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
    对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
    计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
    确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
    根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
    获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
  2. 如权利要求1所述的房产指数数据分析方法,其中,所述计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息,包括:
    计算所述第一目标时序数据对于所述第二目标时序数据的不确定性大小的改变量,得到第一传递熵;
    计算所述第二目标时序数据对于所述第一目标时序数据的不确定性大小的改变量,得到第二传递熵;
    根据所述第一传递熵和所述第二传递熵,计算所述第一目标时序数据和第二目标时序数据之间的目标传递熵;
    基于所述目标传递熵,确定所述第二因子对于所述第一因子的第一领先期限。
  3. 如权利要求1所述的房产指数数据分析方法,其中,所述确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息,包括:
    基于预设的格兰杰因果检验算法,计算所述第一目标时序数据和第二目标时序数据之间的因果关系,并基于所述因果关系确定所述第二因子对于所述第一因子的第二领先期;或者
    基于预设的滞后相关系数算法,计算所述第一目标时序数据和第二目标时序数据之间的多个滞后项系数,并基于所述多个滞后项系数确定所述第二因子对于所述第一因子的第二领先期。
  4. 如权利要求1-3中任一项所述的房产指数数据分析方法,其中,所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据之后,还包括:
    对所述第二目标时序数据进行平稳性检验,得到检验结果;
    若检验结果为第二目标时序数据平稳,则执行所述计算所述第一目标时序数据和第二目标时序数据之间的传递熵的步骤;
    若检验结果为第二目标时序数据不平稳,则通过预设的Barzilai-Borwein梯度算法,确定所述第一因子与第二因子之间的目标领先指标信息。
  5. 如权利要求4所述的房产指数数据分析方法,其中,所述通过预设的Barzilai-Borwein梯度算法,确定所述第一因子与第二因子之间的目标领先指标信息,包括:
    分别对所述第一目标时序数据和第二目标时序数据进行分段线性处理,以确定所述 第一目标时序数据的第一波峰和第一波谷,以及确定所述第二目标时序数据的第二波峰和第二波谷;
    从多个所述第二波峰中确定在所述第一波谷之前的目标波峰,并计算所述目标波峰与所述第一波谷之间的时间差值;
    并从多个所述第二波谷中确定在所述第一波峰之前的所述目标波谷,并计算所述目标波谷与所述第一波峰之间的时间差值;
    根据所述目标波峰与所述第一波谷之间的时间差值、以及所述目标波谷与所述第一波峰之间的时间差值,确定所述第二因子对于所述第一因子的目标领先期限。
  6. 如权利要求1-3中任一项所述的房产指数数据分析方法,其中,所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据,包括:
    通过预设的ARIMA模型分别对所述第一时序数据和第二时序数据进行处理,得到经过调整的所述第一时序数据和所述第二时序数据;
    通过预设的经验模态分解EMD模型分别对经过调整的所述第一时序数据和所述第二时序数据进行处理,得到第一目标时序数据和第二目标时序数据。
  7. 如权利要求6所述的房产指数数据分析方法,其中,所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据之后,还包括:
    分别对所述第一目标时序数据和第二目标时序数据进行滤波处理,得到过滤的第一时序数据和过滤的第二目标时序;
    将所述过滤的第一时序数据作为更新的第一目标时序数据,并将所述过滤的第二目标时序作为更新的第二目标时序数据。
  8. 一种房产指数数据分析装置,其中,所述房产指数数据分析装置包括:
    获取模块,用于获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
    处理模块,用于对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
    第一确定模块,用于计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
    第二确定模块,用于确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
    第三确定模块,用于根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
    预测模块,用于获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
  9. 一种计算机设备,其中,所述计算机设备包括处理器、存储器、以及存储在所述存储器上并可被所述处理器执行的计算机程序,其中所述计算机程序被所述处理器执行时,实现以下步骤:
    获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
    对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
    计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
    确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
    根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
    获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
  10. 如权利要求9所述的计算机设备,其中,所述处理器在实现所述计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息时,用于实现:
    计算所述第一目标时序数据对于所述第二目标时序数据的不确定性大小的改变量,得到第一传递熵;
    计算所述第二目标时序数据对于所述第一目标时序数据的不确定性大小的改变量,得到第二传递熵;
    根据所述第一传递熵和所述第二传递熵,计算所述第一目标时序数据和第二目标时序数据之间的目标传递熵;
    基于所述目标传递熵,确定所述第二因子对于所述第一因子的第一领先期限。
  11. 如权利要求9所述的计算机设备,其中,所述处理器在实现所述确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息时,用于实现:
    基于预设的格兰杰因果检验算法,计算所述第一目标时序数据和第二目标时序数据之间的因果关系,并基于所述因果关系确定所述第二因子对于所述第一因子的第二领先期;或者
    基于预设的滞后相关系数算法,计算所述第一目标时序数据和第二目标时序数据之间的多个滞后项系数,并基于所述多个滞后项系数确定所述第二因子对于所述第一因子的第二领先期。
  12. 如权利要求9-11中任一项所述的计算机设备,其中,所述处理器在实现所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据之后,还用于实现:
    对所述第二目标时序数据进行平稳性检验,得到检验结果;
    若检验结果为第二目标时序数据平稳,则执行所述计算所述第一目标时序数据和第二目标时序数据之间的传递熵的步骤;
    若检验结果为第二目标时序数据不平稳,则通过预设的Barzilai-Borwein梯度算法,确定所述第一因子与第二因子之间的目标领先指标信息。
  13. 如权利要求12所述的计算机设备,其中,所述处理器在实现所述通过预设的Barzilai-Borwein梯度算法,确定所述第一因子与第二因子之间的目标领先指标信息时,用于实现:
    分别对所述第一目标时序数据和第二目标时序数据进行分段线性处理,以确定所述第一目标时序数据的第一波峰和第一波谷,以及确定所述第二目标时序数据的第二波峰和第二波谷;
    从多个所述第二波峰中确定在所述第一波谷之前的目标波峰,并计算所述目标波峰与所述第一波谷之间的时间差值;
    并从多个所述第二波谷中确定在所述第一波峰之前的所述目标波谷,并计算所述目标波谷与所述第一波峰之间的时间差值;
    根据所述目标波峰与所述第一波谷之间的时间差值、以及所述目标波谷与所述第一波峰之间的时间差值,确定所述第二因子对于所述第一因子的目标领先期限。
  14. 如权利要求9-11中任一项所述的计算机设备,其中,所述处理器在实现所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据时,用于实现:
    通过预设的ARIMA模型分别对所述第一时序数据和第二时序数据进行处理,得到经过调整的所述第一时序数据和所述第二时序数据;
    通过预设的经验模态分解EMD模型分别对经过调整的所述第一时序数据和所述第二时序数据进行处理,得到第一目标时序数据和第二目标时序数据。
  15. 一种计算机可读存储介质,其中,所述计算机可读存储介质上存储有计算机程序,其中所述计算机程序被处理器执行时,实现以下步骤:
    获取第一因子的第一时序数据和第二因子的第二时序数据,所述第一因子为符合房地产业要素的房产指数因子,所述第二因子为影响房地产业发展的外部环境因子;
    对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据;
    计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息;
    确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息;
    根据所述第一领先指标信息和所述第二领先指标信息,确定所述第一因子与第二因子之间的目标领先指标信息;
    获取所述第二因子的当前时序数据,并基于所述目标领先指标信息和所述当前时序数据,预测所述第一因子的目标时序数据。
  16. 如权利要求15所述的计算机可读存储介质,其中,所述处理器在实现所述计算所述第一目标时序数据和第二目标时序数据之间的传递熵,并基于所述传递熵确定所述第一因子与第二因子之间的第一领先指标信息时,用于实现:
    计算所述第一目标时序数据对于所述第二目标时序数据的不确定性大小的改变量,得到第一传递熵;
    计算所述第二目标时序数据对于所述第一目标时序数据的不确定性大小的改变量,得到第二传递熵;
    根据所述第一传递熵和所述第二传递熵,计算所述第一目标时序数据和第二目标时序数据之间的目标传递熵;
    基于所述目标传递熵,确定所述第二因子对于所述第一因子的第一领先期限。
  17. 如权利要求15所述的计算机可读存储介质,其中,所述处理器在实现所述确定所述第一目标时序数据和第二目标时序数据之间的线性关系,并基于所述线性关系确定所述第一因子与第二因子之间的第二领先指标信息时,用于实现:
    基于预设的格兰杰因果检验算法,计算所述第一目标时序数据和第二目标时序数据之间的因果关系,并基于所述因果关系确定所述第二因子对于所述第一因子的第二领先期;或者
    基于预设的滞后相关系数算法,计算所述第一目标时序数据和第二目标时序数据之间的多个滞后项系数,并基于所述多个滞后项系数确定所述第二因子对于所述第一因子的第二领先期。
  18. 如权利要求15-17中任一项所述的计算机可读存储介质,其中,所述处理器在实现所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据之后,还用于实现:
    对所述第二目标时序数据进行平稳性检验,得到检验结果;
    若检验结果为第二目标时序数据平稳,则执行所述计算所述第一目标时序数据和第 二目标时序数据之间的传递熵的步骤;
    若检验结果为第二目标时序数据不平稳,则通过预设的Barzilai-Borwein梯度算法,确定所述第一因子与第二因子之间的目标领先指标信息。
  19. 如权利要求18所述的计算机可读存储介质,其中,所述处理器在实现所述通过预设的Barzilai-Borwein梯度算法,确定所述第一因子与第二因子之间的目标领先指标信息时,用于实现:
    分别对所述第一目标时序数据和第二目标时序数据进行分段线性处理,以确定所述第一目标时序数据的第一波峰和第一波谷,以及确定所述第二目标时序数据的第二波峰和第二波谷;
    从多个所述第二波峰中确定在所述第一波谷之前的目标波峰,并计算所述目标波峰与所述第一波谷之间的时间差值;
    并从多个所述第二波谷中确定在所述第一波峰之前的所述目标波谷,并计算所述目标波谷与所述第一波峰之间的时间差值;
    根据所述目标波峰与所述第一波谷之间的时间差值、以及所述目标波谷与所述第一波峰之间的时间差值,确定所述第二因子对于所述第一因子的目标领先期限。
  20. 如权利要求15-17中任一项所述的计算机可读存储介质,其中,所述处理器在实现所述对所述第一时序数据和第二时序数据均进行季节性调整和平稳化处理,得到第一目标时序数据和第二目标时序数据时,用于实现:
    通过预设的ARIMA模型分别对所述第一时序数据和第二时序数据进行处理,得到经过调整的所述第一时序数据和所述第二时序数据;
    通过预设的经验模态分解EMD模型分别对经过调整的所述第一时序数据和所述第二时序数据进行处理,得到第一目标时序数据和第二目标时序数据。
PCT/CN2021/097416 2021-04-30 2021-05-31 房产指数数据分析方法、装置、设备及存储介质 WO2022227219A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110486524.6A CN113205257A (zh) 2021-04-30 2021-04-30 房产指数数据分析方法、装置、设备及存储介质
CN202110486524.6 2021-04-30

Publications (1)

Publication Number Publication Date
WO2022227219A1 true WO2022227219A1 (zh) 2022-11-03

Family

ID=77028589

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/097416 WO2022227219A1 (zh) 2021-04-30 2021-05-31 房产指数数据分析方法、装置、设备及存储介质

Country Status (2)

Country Link
CN (1) CN113205257A (zh)
WO (1) WO2022227219A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116049341B (zh) * 2023-03-08 2023-08-15 北京七兆科技有限公司 一种水文数据标准化方法、装置、设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105934754A (zh) * 2014-02-14 2016-09-07 欧姆龙株式会社 因果网络生成系统和因果关系的数据结构
US10055788B1 (en) * 2009-11-18 2018-08-21 Federal Home Loan Morgage Corporation Systems, methods, and computer-readable storage media for calculating a housing volatility index
CN109635996A (zh) * 2018-10-31 2019-04-16 平安直通咨询有限公司 房价走势前瞻模型构建方法、装置及存储介质
CN111598334A (zh) * 2020-05-14 2020-08-28 深圳禾略贝塔信息科技有限公司 一种地产行业周期识别方法、装置、系统、终端以及存储介质
CN112348222A (zh) * 2020-05-08 2021-02-09 东南大学 一种基于因果逻辑和图卷积特征提取的网络耦合时序信息流预测方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10055788B1 (en) * 2009-11-18 2018-08-21 Federal Home Loan Morgage Corporation Systems, methods, and computer-readable storage media for calculating a housing volatility index
CN105934754A (zh) * 2014-02-14 2016-09-07 欧姆龙株式会社 因果网络生成系统和因果关系的数据结构
CN109635996A (zh) * 2018-10-31 2019-04-16 平安直通咨询有限公司 房价走势前瞻模型构建方法、装置及存储介质
CN112348222A (zh) * 2020-05-08 2021-02-09 东南大学 一种基于因果逻辑和图卷积特征提取的网络耦合时序信息流预测方法
CN111598334A (zh) * 2020-05-14 2020-08-28 深圳禾略贝塔信息科技有限公司 一种地产行业周期识别方法、装置、系统、终端以及存储介质

Also Published As

Publication number Publication date
CN113205257A (zh) 2021-08-03

Similar Documents

Publication Publication Date Title
Guarniero et al. The iterated auxiliary particle filter
Corlay et al. Multifractional stochastic volatility models
WO2020119051A1 (zh) 云平台资源使用预测方法及终端设备
US9928320B2 (en) Techniques for estimating compound probability distribution by simulating large empirical samples with scalable parallel and distributed processing
Møller et al. Probabilistic forecasts of wind power generation by stochastic differential equation models
WO2013114385A1 (en) System and method for generating marketing- mix solution
Wadsworth Exploiting structure of maximum likelihood estimators for extreme value threshold selection
Schneider et al. Threshold selection in univariate extreme value analysis
CN112132485A (zh) 一种指标数据处理方法、装置、电子设备及存储介质
CN111524001B (zh) 用户授信额度的预测方法、装置和相关设备
Eberlein et al. Variational solutions of the pricing PIDEs for European options in Lévy models
CN111415192A (zh) 基于大数据的用户用水性质预测方法
CN112907064A (zh) 基于自适应窗的电量预测方法、装置、存储介质及终端
WO2022227219A1 (zh) 房产指数数据分析方法、装置、设备及存储介质
CN115564152A (zh) 基于stirpat模型的碳排放量预测方法及装置
CN110766184A (zh) 订单量预测方法和装置
Guan Bernstein polynomial model for grouped continuous data
CN114118570A (zh) 业务数据预测方法及装置、电子设备和存储介质
Perera et al. A goodness-of-fit test for a class of autoregressive conditional duration models
Das et al. Functional regression-based monitoring of quality of service in hospital emergency departments
Almomani et al. Selecting a good stochastic system for the large number of alternatives
Andersson et al. Mortality forecasting using a Lexis-based state-space model
Figueroa‐López Central Limit Theorems for the Non‐Parametric Estimation of Time‐Changed Lévy Models
Laurini et al. Regular variation and extremal dependence of garch residuals with application to market risk measures
US20150199695A1 (en) Reporting on Technology Sector Sizes Using Patent Assets

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21938689

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21938689

Country of ref document: EP

Kind code of ref document: A1