WO2021004324A1 - 资源数据的处理方法、装置、计算机设备和存储介质 - Google Patents

资源数据的处理方法、装置、计算机设备和存储介质 Download PDF

Info

Publication number
WO2021004324A1
WO2021004324A1 PCT/CN2020/098879 CN2020098879W WO2021004324A1 WO 2021004324 A1 WO2021004324 A1 WO 2021004324A1 CN 2020098879 W CN2020098879 W CN 2020098879W WO 2021004324 A1 WO2021004324 A1 WO 2021004324A1
Authority
WO
WIPO (PCT)
Prior art keywords
resource data
data
index
prediction
trained
Prior art date
Application number
PCT/CN2020/098879
Other languages
English (en)
French (fr)
Inventor
刘媛源
郑子欧
张翔
于修铭
汪伟
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 WO2021004324A1 publication Critical patent/WO2021004324A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/957Browsing optimisation, e.g. caching or content distillation
    • 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
    • G06Q10/04Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • 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
    • G06Q10/0639Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393Score-carding, benchmarking or key performance indicator [KPI] analysis

Definitions

  • This application relates to the technical field of predictive analysis, in particular to a method, device, computer equipment and storage medium for processing resource data.
  • Resource data forecast information refers to the proportion of resource data such as interest rates in a period of time in the future; different resource data types correspond to different resource data forecast information; in order to understand the proportion of resource data changes in a period of time in the future, forecast the resource data It is very important.
  • the current forecast of the proportion of resource data changes is generally through a single analysis and calculation of a large amount of manual collected specific data, such as economic data, capital data, and policy data, through the server, to initially predict the changes in the resource data in the future. trend.
  • the inventor realizes that the changing trend of resource data is affected by multiple factors, and the main influencing factors corresponding to different resource data types are different, resulting in many data collected manually without actual reference value; and the existing server resources are limited.
  • the server needs to perform calculation and evaluation based on a large amount of specific data collected manually, resulting in a long calculation cycle, resulting in a waste of server resources, and a decrease in server resource utilization.
  • a resource data processing method, device, computer equipment, and storage medium are provided.
  • a method for processing resource data includes:
  • the query request is used to obtain prediction information of the resource data to be predicted, and the query request carries the resource data type of the resource data to be predicted;
  • a resource data processing device includes:
  • the request receiving module is configured to receive a query request sent by a terminal; the query request is used to obtain prediction information of resource data to be predicted, and the query request carries the resource data type of the resource data to be predicted;
  • a data acquisition module configured to acquire a predictive index corresponding to the resource data type, and obtain data to be predicted corresponding to the predictive index from a database;
  • the data extraction module is used to obtain the target data identifier of the data to be predicted, and extract corresponding target data from the data to be predicted according to the target data identifier, as the target data corresponding to the predictive index ;
  • a data conversion module configured to obtain a preset data conversion instruction, and according to the preset data conversion instruction, convert target data corresponding to the predictive index to obtain a reference value corresponding to the predictive index;
  • the change ratio obtaining module is used to input the obtained reference value corresponding to the predictive index into a pre-trained resource data prediction model to obtain the resource data change ratio corresponding to the resource data type;
  • the information generating module is configured to generate resource data prediction information according to the resource data change ratio corresponding to the resource data type, and send the resource data prediction information to the terminal for display.
  • a computer device including a memory and one or more processors, the memory stores computer readable instructions, when the computer readable instructions are executed by the processor, the one or more processors execute The following steps:
  • the query request is used to obtain prediction information of the resource data to be predicted, and the query request carries the resource data type of the resource data to be predicted;
  • One or more computer-readable storage media storing computer-readable instructions.
  • the one or more processors execute the following steps:
  • the query request is used to obtain prediction information of the resource data to be predicted, and the query request carries the resource data type of the resource data to be predicted;
  • the server uses a pre-trained resource data prediction model to evaluate and analyze only the data to be predicted corresponding to the predictive index related to the resource data change ratio of the resource data type. For each resource data type, a large amount of specific data collected manually is calculated and evaluated to avoid waste of server resources, thereby improving the resource utilization of the server; at the same time, combined with the pre-trained resource data prediction model, the resource data type The analysis of the relevant data to be predicted realizes the purpose of comprehensive tracking and evaluation of the resource data change ratio corresponding to the resource data type, and further improves the accuracy of the obtained resource data prediction information.
  • Fig. 1 is an application scenario diagram of a resource data processing method according to one or more embodiments
  • FIG. 2 is a schematic flowchart of a method for processing resource data according to one or more embodiments
  • FIG. 3 is a schematic flowchart of steps for obtaining a resource data change ratio corresponding to a resource data type according to one or more embodiments
  • Fig. 4 is a block diagram of an apparatus for processing resource data according to one or more embodiments
  • Figure 5 is a block diagram of a computer device according to one or more embodiments.
  • the resource data processing method provided in this application can be applied to the application environment shown in FIG. 1.
  • the terminal 110 and the server 120 communicate through the network.
  • the terminal 110 is installed with an application program, through which the user can query the resource data change ratio corresponding to the resource data type, such as the interest rate fluctuation ratio corresponding to the interest rate type.
  • the terminal 110 responds to the user's input operation on the query interface displayed by the application program, generates a query request for obtaining resource data prediction information, and sends the query request to the server 120.
  • the server 120 parses the query request to obtain the resource data type of the resource data to be predicted input by the user; obtains the predictive index corresponding to the resource data type, queries the database according to the resource data type, and obtains the to-be-predicted data corresponding to the predictive index from the database; According to the target data identifier of the data to be predicted, the corresponding target data is extracted from the data to be predicted according to the target data identifier as the target data corresponding to the predictive index; the preset data conversion instruction is obtained, and the preset data conversion instruction is obtained, Transform the target data corresponding to the predictive index to obtain the reference value corresponding to the predictive index; input the obtained reference value corresponding to each predictive index into the pre-trained resource data prediction model to obtain the resource data change corresponding to the resource data type Proportion: Generate resource data prediction information according to the resource data change ratio corresponding to the resource data type, and send the resource data prediction information to the terminal 110.
  • the terminal 110 displays the resource data change ratio corresponding to the resource data type of the resource data input by the user according to the received resource data prediction information.
  • the terminal 110 may be, but is not limited to, various personal computers, notebook computers, smart phones, and tablet computers, and the server 120 may be implemented by an independent server or a server cluster composed of multiple servers.
  • a method for processing resource data is provided. Taking the method applied to the server in FIG. 1 as an example for description, the method includes the following steps:
  • Step S201 Receive a query request sent by a terminal; the query request is used to obtain prediction information of the resource data to be predicted, and the query request carries the resource data type of the resource data to be predicted.
  • the resource data is the exponential value gain ratio, which represents the percentage of the numerical gain, such as interest rate;
  • the resource data prediction information refers to the proportion of the resource data to be predicted in the future, such as interest rate prediction information;
  • the resource data type is Refers to the information used to identify resource data to be predicted, such as loan interest rates, deposit interest rates, etc.
  • the terminal installs an application program for querying resource data prediction information corresponding to the resource data type, and the user can query the resource data prediction information corresponding to the resource data type by logging into the application program.
  • the terminal responds to the user's input operation on the query interface displayed by the application program, generates a query request for obtaining resource data prediction information, and sends the query request to the server.
  • the server parses the received query request to obtain the resource data type of the resource data to be predicted.
  • the user can also enter the resource data prediction information query interface by logging into the browser running on the terminal. The user inputs the resource data type of the resource data to be predicted based on the query interface, and sends the query request carrying the resource data type to the server through the terminal.
  • Step S202 Obtain the predictive index corresponding to the resource data type, and obtain the data to be predicted corresponding to the predictive index from the database.
  • predictive indicators refer to indicators that affect the proportion of resource data changes, such as news, policies, money supply, total social financing, expert research reports, etc.; different resource data types correspond to different predictive indicators.
  • the server uses predictive indicators as search keywords, crawls data corresponding to each predictive index from the Internet in advance, and performs preprocessing operations on the crawled data corresponding to each predictive index, such as removing noise and filtering interference information Etc., obtain the key data content corresponding to each predictive index; use the obtained key data content as the data to be predicted, and store it in a pre-established database according to the predictive index classification; it is convenient for the subsequent server to query the database to obtain the corresponding predictive index The data to be predicted.
  • the server re-obtains the to-be-predicted data corresponding to each predictive index based on the Internet.
  • the re-acquired to-be-predicted data is used to overwrite the original data corresponding to the predictive index in the database , Updating the data in the database in a timely manner is conducive to improving the accuracy and timeliness of the obtained data to be predicted corresponding to each index, and further improving the accuracy of the subsequent resource data prediction information.
  • the server filters out the known resource data types that match the acquired resource data types; obtains the predictive index corresponding to the known resource data types, which will match the known resource data types
  • the corresponding predictive index is used as the predictive index corresponding to the acquired resource data type;
  • the database is queried according to the predictive index corresponding to the acquired resource data type, and the data to be predicted corresponding to the predictive index is obtained from the database to facilitate subsequent servers according to
  • the to-be-predicted data corresponding to the predictive index determines the resource data change ratio corresponding to the resource data type, which realizes a comprehensive tracking evaluation of the resource data change ratio, thereby improving the accuracy of the obtained resource data prediction information.
  • Step S203 Obtain the target data identifier of the data to be predicted, and extract the corresponding target data from the data to be predicted according to the target data identifier, as the target data corresponding to the predictive index.
  • the target data identifier is used to identify the key information in the data to be predicted, and the target data identifiers corresponding to different target data are different; the target data refers to the key information in the data to be predicted; suppose the data to be predicted is news Data, then the target data in the data to be predicted refers to the key information in the news data.
  • the server obtains the target data identifier of the data to be predicted in advance, extracts the target data corresponding to the target data identifier from the data to be predicted, and recognizes the target data as the target data corresponding to the predictive index. In this way, it is beneficial to extract the target data in the data to be predicted, avoid unnecessary data interference, and further reduce the data that the server needs to process, thereby improving the resource utilization rate of the server.
  • step S204 a preset data conversion instruction is acquired, and the target data corresponding to the predictive index is converted according to the preset data conversion instruction to obtain a reference value corresponding to the predictive index.
  • the data conversion instruction is an instruction that can convert the to-be-predicted data corresponding to the predictive index into the corresponding reference value; different predictive indexes have different corresponding data conversion instructions.
  • the reference value refers to the value converted from the target data corresponding to the predictive index based on the data conversion instruction; it is used to convert the target data into a value that is convenient for the calculation of the prediction model. For example, the target data corresponding to the total social financing index is transformed into the corresponding total social financing level, then the total social financing level is the reference value corresponding to the total social financing index.
  • the server obtains the preset data conversion instruction, converts the target data corresponding to the predictive index according to the preset data conversion instruction, obtains the corresponding data level, and recognizes the obtained data level as a reference corresponding to the predictive index Value; it is convenient for the subsequent server to predict the proportion of resource data change based on the obtained reference value, avoiding redundant data interference, thereby improving the accuracy of the obtained resource data prediction information.
  • the target data corresponding to the total social financing index is converted into the corresponding total social financing level, and the obtained total social financing level is used as the reference value corresponding to the total social financing index; If the total amount of social financing is c, the corresponding total level of social financing is level 1, that is, the reference value corresponding to the total social financing index is 1; assuming that the total amount of social financing is d, the corresponding total level of social financing is The second level, that is, the reference value corresponding to the total social financing index is 2.
  • the target data corresponding to the policy indicator For theoretical target data, for example, according to the degree of influence of the target data corresponding to the policy indicator on the increase in the proportion of resource data, convert the target data into the corresponding policy level, and use the obtained policy level as the reference value corresponding to the policy indicator; For example, the target data corresponding to the policy indicator has a higher degree of influence on the increase in the proportion of resource data change, and the corresponding policy level is level 1, that is, the reference value corresponding to the policy indicator is 1.
  • Step S205 Input the obtained data to be predicted corresponding to the predictive index into a pre-trained resource data prediction model to obtain a resource data change ratio corresponding to the resource data type.
  • the resource data prediction model is a model that can determine the resource data change ratio corresponding to the resource data type according to the input reference value corresponding to the prediction index.
  • the resource data change ratio is the change ratio of the exponential value gain ratio in the future period of time, such as the interest rate change ratio, which is used to indicate the fluctuation trend of the numerical gain ratio in the future period of time.
  • the server inputs the obtained reference value corresponding to each predictive index into the pre-trained resource data prediction model, and calculates and analyzes the reference value corresponding to each predictive index through the resource data predictive model to obtain the resource data prediction result corresponding to each predictive index ; Integrate the resource data prediction results corresponding to each predictive index to obtain the resource data change ratio corresponding to the resource data type, thereby improving the accuracy of the resource data prediction information obtained.
  • Step S206 Generate resource data prediction information according to the resource data change ratio corresponding to the resource data type, and send the resource data prediction information to the terminal for display.
  • the server obtains the preset resource data prediction information template, imports the resource data change ratio corresponding to the resource data type into the preset resource data prediction information template, generates corresponding resource data prediction information, and combines the generated resource
  • the data prediction information is sent to the corresponding terminal, and the resource data change ratio corresponding to the resource data type is displayed through the display interface of the terminal, which is convenient for users to intuitively and comprehensively understand the resource data change ratio of the resource data type in the future.
  • the user conducts a comprehensive tracking evaluation of the resource data change ratio of the resource data type.
  • the server uses a pre-trained resource data prediction model to evaluate and analyze only the data to be predicted corresponding to the predictive index related to the resource data change ratio of the resource data type, without requiring for each resource data type , All perform calculation and evaluation on a large amount of specific data collected manually, avoiding the waste of server resources, thereby improving the resource utilization of the server; at the same time, combining the pre-trained resource data prediction model to analyze the data to be predicted related to the resource data type , To achieve the purpose of comprehensive tracking and evaluation of the resource data change ratio corresponding to the resource data type, and further improve the accuracy of the obtained resource data prediction information.
  • the step of obtaining the predictive index corresponding to the resource data type specifically includes: matching the resource data type with a preset known resource data type; if the resource data type matches the known resource data type Type matching, obtain the predictive index corresponding to the known resource data type; use the obtained predictive index corresponding to the known resource data type as the predictive index corresponding to the resource data type.
  • the server respectively calculates the matching degree between the resource data type and each preset known resource data type; from the preset known resource data types, the known resource data type with the largest matching degree is filtered out, The known resource data type with the greatest matching degree is taken as the known resource data type that successfully matches the obtained resource data type; the predictive index corresponding to the known resource data type with the greatest matching degree is obtained, and the known resource data will be matched
  • the predictive index corresponding to the type is used as the predictive index corresponding to the acquired resource data type; it is convenient for the subsequent server to query the database to obtain the data to be predicted corresponding to each predictive index.
  • the step S204 described above is to obtain a preset data conversion instruction, and according to the preset data conversion instruction, convert the target data corresponding to the predictive index to obtain the reference value corresponding to the predictive index, including: extracting the preset data The data conversion rule in the data conversion instruction; the data conversion rule is the conversion rule between the target data and the reference value; according to the data conversion rule, the target data corresponding to the predictive index is converted to obtain the reference value corresponding to the predictive index.
  • the data conversion rule refers to a rule for converting target data corresponding to a predictive indicator into a corresponding reference value; different predictive indicators have different corresponding data conversion rules.
  • the server obtains the identifier of the data conversion rule, and extracts the corresponding data conversion rule from the data conversion instruction according to the identifier of the data conversion rule; according to the data conversion rule, transforms the target data corresponding to the predictive index to obtain the corresponding data Level, and identify the obtained data level as the reference value corresponding to the predictive index.
  • the server can also analyze the corresponding reference value through the index prediction model corresponding to each prediction index, and integrate the resources of each index prediction model The data change sub-proportion to improve the accuracy of the obtained resource data prediction information.
  • the step of inputting the obtained reference value corresponding to the predictive index into the pre-trained resource data prediction model, and obtaining the resource data change ratio corresponding to the resource data type specifically includes:
  • step S301 the reference value corresponding to each predictive index is input into the corresponding index prediction model to obtain the resource data change sub-proportion corresponding to each predictive index.
  • Step S302 Obtain preset weighting factors corresponding to each index prediction model.
  • Step S303 Perform a weighted calculation on the corresponding resource data change sub-proportions according to the weighting factors corresponding to the respective index prediction models to obtain the resource data change proportions corresponding to the resource data types.
  • the index prediction model refers to a model that can roughly determine the corresponding resource data change ratio based on the input reference value corresponding to the prediction index.
  • the policy indicator forecasting model can roughly predict the sub-proportion of changes in resource data based on the reference values of major events, important policies and other policy-related data to be predicted; for example, the Sino-US trade war will affect the exchange rate of RMB against the US dollar, and the exchange rate will affect interest rates. The change trend in the future; it needs to be explained that other index prediction models are also implemented based on the same principle, so I won't repeat them here.
  • the sub-proportion of resource data change refers to the proportion of resource data change roughly predicted based on the reference value corresponding to the data to be predicted corresponding to a single predictive indicator. It represents the change proportion of the numerical gain ratio in the future under a single predictive indicator.
  • the resource data change ratio refers to the total change ratio predicted by integrating the resource data change sub-proportions corresponding to multiple predictive indicators.
  • the weight factor refers to the degree of influence of the resource data change sub-proportion output by each index prediction model on the final resource data change ratio; the larger the weight factor, the greater the influence on the final resource data change ratio.
  • the server inputs the reference value corresponding to each predictive index into the corresponding index prediction model, calculates and evaluates the corresponding reference value through the index prediction model, determines the corresponding resource data change ratio, and uses the determined resource data change ratio as The resource data change sub-proportion corresponding to the predicted index; combined with preset weight factors corresponding to each indicator prediction model, weighted calculation is performed on the corresponding resource data change sub-proportion to obtain the resource data change ratio corresponding to the resource data type.
  • the predictive indicators are a, b, c, d
  • the resource data change sub-proportions corresponding to the predictive indicators a, b, c, d are y1, y2, y3,, y4, and the predictive indicators a, b, c, d
  • the reference value corresponding to each predictive index is analyzed through each index prediction model, avoiding redundant data interference, reducing the data that the server needs to process, and improving the resource utilization rate of the server; Analyze the resource data change ratio from the perspective, thereby achieving the purpose of comprehensive tracking and evaluating the resource data change ratio, avoiding errors, and improving the accuracy of the resource data prediction information obtained.
  • the above step S205 generating resource data prediction information according to the resource data change ratio corresponding to the resource data type, and sending the resource data prediction information to the terminal for display, includes: obtaining the preset resource data of each predictive index The corresponding import location in the map template; import the to-be-predicted data and resource data change sub-proportions corresponding to each predictive index into the corresponding import location in the preset resource data map template to generate a resource data map; based on the resource data map and The resource data change ratio corresponding to the resource data type generates resource data reference information; the resource data map, the resource data change ratio corresponding to the resource data type, and the resource data reference information are sequentially imported into the preset resource data prediction information template to generate the corresponding And send the resource data prediction information to the terminal for display.
  • resource data reference information refers to information summarized based on the generated resource data map and the resource data change ratio corresponding to the resource data type; for example, which predictive indicators have a greater impact on the resource data change ratio, and what needs to be taken Countermeasures, etc.; it is convenient for users to view directly without manual summary.
  • the server imports the generated resource data map, the resource data change ratio corresponding to the resource data type, and the resource data reference information in turn into the preset resource data prediction information template, generates corresponding resource data prediction information, and
  • the resource data prediction information is sent to the terminal, and the resource data map, the resource data change ratio corresponding to the resource data type, and the resource data reference information are displayed through the display interface of the terminal; for example, the display interface of the terminal displays the interest rate map and the corresponding interest rate type.
  • Interest rate change ratio and interest rate reference information Through this embodiment, it is convenient for the user to intuitively and comprehensively understand the change ratio of resource data in a period of time in the future, which is beneficial for the user to comprehensively track and evaluate the resource data change ratio corresponding to the resource data type.
  • the server may also detect the confirmation information of the resource data prediction information returned by the terminal; Within the preset time range, if the confirmation information of the resource data prediction information returned by the terminal is detected, the resource data prediction information is marked as sent; if within the preset time range, the confirmation of the resource data prediction information returned by the terminal is not detected Information, the resource data prediction information is re-sent to the corresponding terminal.
  • the index prediction model can be trained multiple times.
  • the index prediction model is obtained by the following method: respectively obtaining sample prediction data corresponding to each index prediction model to be trained; training the corresponding index prediction model to be trained according to the sample prediction data to obtain training Obtain the prediction error between the resource data change sub-proportion output by the trained indicator prediction model and the corresponding actual resource data change sub-proportion; when the prediction error is greater than or equal to the corresponding preset threshold, according to the prediction Error
  • the index prediction model to be trained is repeatedly trained until the prediction error obtained according to the trained index prediction model is less than the corresponding preset threshold.
  • the server adjusts the parameters of the index prediction model according to the prediction error; re-evaluates the sample prediction data according to the adjusted index prediction model, and obtains the result of the adjusted index prediction model
  • the prediction error between the resource data change sub-proportion and the corresponding actual resource data change sub-proportion the parameters of the indicator prediction model are adjusted again according to the prediction error to retrain the indicator prediction model until it is predicted according to the trained indicator
  • the prediction error between the resource data change sub-proportion obtained by the model and the corresponding actual resource data change sub-proportion is less than the corresponding preset threshold; when the prediction error obtained by each index prediction model is less than the corresponding preset threshold, it is obtained separately
  • the current index prediction models; the current index prediction models are used as the trained index prediction models.
  • the server trains each index prediction model multiple times according to the prediction error, which is beneficial to output more accurate resource data change sub-proportions through each index prediction model, thereby improving the accuracy of resource data prediction of each index prediction model.
  • Rate which helps to improve the accuracy of the resource data prediction information obtained subsequently.
  • the server may also train the resource data prediction model multiple times to improve the accuracy of the obtained resource data prediction information.
  • the resource data prediction model is obtained by the following method: respectively obtaining sample prediction data and weighting factors corresponding to each trained index prediction model; according to each trained index prediction model and corresponding sample prediction data And the weighting factor, the resource data prediction model to be trained is trained to obtain the resource data prediction model after training; the first value between the resource data change ratio output by the trained resource data prediction model and the corresponding actual resource data change ratio is obtained Prediction error; when the first prediction error is greater than or equal to the first preset threshold, the weighting factor corresponding to each trained index prediction model is adjusted according to the first prediction error, and the resource data to be trained according to the adjusted weighting factor The prediction model is repeatedly trained until the first prediction error obtained from the trained resource data prediction model is less than the first preset threshold.
  • the server adjusts the weighting factor corresponding to each trained index prediction model according to the first prediction error, and according to the adjusted weighting factor, the resource data to be trained
  • the prediction model is retrained; the first prediction error between the resource data change ratio obtained according to the resource data prediction model after the retraining and the corresponding actual resource data change ratio is obtained, and each trained index is predicted according to the first prediction error
  • the weighting factor corresponding to the model is adjusted again to retrain the resource data prediction model to be trained until the first prediction between the resource data change ratio obtained by the trained resource data prediction model and the corresponding actual resource data change ratio
  • the error is less than the first preset threshold; when the obtained first prediction error is less than the first preset threshold, obtain the current resource data prediction model as the trained resource data prediction model; and use the current weighting factors as The preset weight factors corresponding to each trained index prediction model.
  • by continuously adjusting the weight factors of each trained index prediction model to train the resource data prediction model multiple times by continuously adjusting the weight factors of each trained index prediction model to train the
  • a device for processing resource data including: a request receiving module 410, a data acquisition module 420, a data extraction module 430, a data conversion module 440, and a change ratio acquisition module 450 And information generating module 460, where:
  • the request receiving module 410 is configured to receive a query request sent by the terminal; the query request is used to obtain prediction information of the resource data to be predicted, and the query request carries the resource data type of the resource data to be predicted.
  • the data acquisition module 420 is configured to acquire the predictive index corresponding to the resource data type, and obtain the data to be predicted corresponding to the predictive index from the database.
  • the data extraction module 430 is configured to obtain the target data identifier of the data to be predicted, and extract the corresponding target data from the data to be predicted according to the target data identifier, as the target data corresponding to the predictive index.
  • the data conversion module 440 is configured to obtain a preset data conversion instruction, and according to the preset data conversion instruction, convert the target data corresponding to the predictive index to obtain a reference value corresponding to the predictive index.
  • the change ratio obtaining module 450 is configured to input the obtained reference values corresponding to each predictive index into the pre-trained resource data prediction model to obtain the resource data change ratio corresponding to the resource data type.
  • the information generating module 460 is configured to generate resource data prediction information according to the resource data change ratio corresponding to the resource data type, and send the resource data prediction information to the terminal for display.
  • the data acquisition module is also used to match the resource data type with a preset known resource data type; if the resource data type matches the known resource data type, acquire the data type corresponding to the known resource data type. Predictive index; the obtained predictive index corresponding to the known resource data type is used as the predictive index corresponding to the resource data type.
  • the data conversion module is also used to extract the data conversion rule in the preset data conversion instruction; the data conversion rule is the conversion rule between the target data and the reference value; according to the data conversion rule, the pair corresponds to the predictive index Transform the target data to obtain the reference value corresponding to the predictive index.
  • the change ratio obtaining module is also used to input the reference value corresponding to each predictive index into the corresponding index prediction model to obtain the resource data change sub-ratio corresponding to each predictive index; to obtain the preset and each The weighting factor corresponding to the indicator prediction model; respectively, according to the weighting factor corresponding to each indicator prediction model, the corresponding resource data change sub-proportions are weighted and calculated to obtain the resource data change ratio corresponding to the resource data type.
  • the information generation module is also used to obtain the corresponding import position of each predictive index in the preset resource data map template; import the to-be-predicted data and resource data change sub-proportion corresponding to each predictive index into the predictive index.
  • Set the corresponding import location in the resource data map template to generate a resource data map; generate resource data reference information according to the resource data map and the resource data change ratio corresponding to the resource data type; combine the resource data map and the resource data corresponding to the resource data type
  • the change ratio and the resource data reference information are sequentially imported into the preset resource data prediction information template, the corresponding resource data prediction information is generated, and the resource data prediction information is sent to the terminal for display.
  • the resource data processing device further includes an index prediction model training module, which is used to obtain sample prediction data corresponding to each index prediction model to be trained; predict the corresponding index to be trained based on the sample prediction data
  • the model is trained to obtain the index prediction model after training; the prediction error between the resource data change sub-ratio output by the trained index prediction model and the corresponding actual resource data change sub-ratio is obtained; when the prediction error is greater than or equal to the corresponding prediction
  • the index prediction model to be trained is repeatedly trained according to the prediction error until the prediction error obtained according to the trained index prediction model is less than the corresponding preset threshold.
  • the resource data processing device further includes a resource data model training module, which is used to obtain sample prediction data and weighting factors corresponding to each trained index prediction model; according to each trained index prediction model and Corresponding sample prediction data and weight factors, train the resource data prediction model to be trained to obtain the resource data prediction model after training; obtain the resource data change ratio output by the resource data prediction model after training and the corresponding actual resource data change ratio
  • a resource data model training module which is used to obtain sample prediction data and weighting factors corresponding to each trained index prediction model; according to each trained index prediction model and Corresponding sample prediction data and weight factors, train the resource data prediction model to be trained to obtain the resource data prediction model after training; obtain the resource data change ratio output by the resource data prediction model after training and the corresponding actual resource data change ratio
  • the resource data prediction model to be trained is repeatedly trained until the first prediction error obtained according to the trained resource data prediction model is less than the first preset threshold.
  • the resource data processing device uses a pre-trained resource data prediction model to evaluate and analyze only the data to be predicted corresponding to the predictive index related to the resource data change ratio of the resource data type, and does not need to target each type of resource data Types are calculated and evaluated on a large amount of specific data collected manually to avoid waste of server resources, thereby improving the resource utilization of the server; at the same time, combined with the pre-trained resource data prediction model, the data to be predicted related to the resource data type is performed
  • the analysis achieves the purpose of comprehensive tracking and evaluation of the resource data change ratio corresponding to the resource data type, and further improves the accuracy of the resource data prediction information obtained.
  • Each module in the above-mentioned resource data processing device can be implemented in whole or in part by software, hardware, and a combination thereof.
  • the foregoing modules may be embedded in the form of hardware or independent of the processor in the computer device, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the foregoing modules.
  • a computer device is provided.
  • the computer device may be a server, and its internal structure diagram may be as shown in FIG. 5.
  • the computer equipment includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide calculation and control capabilities.
  • the memory of the computer device includes a non-volatile or volatile storage medium and internal memory.
  • the non-volatile or volatile storage medium stores an operating system, computer readable instructions, and a database.
  • the internal memory provides an environment for the operation of the operating system and computer-readable instructions in the non-volatile or volatile storage medium.
  • the database of the computer equipment is used to store the data to be evaluated corresponding to each evaluation index.
  • the network interface of the computer device is used to communicate with an external terminal through a network connection.
  • the computer-readable instructions are executed by the processor to implement a method for processing resource data.
  • FIG. 5 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied.
  • the specific computer device may Including more or fewer parts than shown in the figure, or combining some parts, or having a different arrangement of parts.
  • a computer device including a memory and one or more processors, in which computer readable instructions are stored, and when the computer readable instructions are executed by the processor, the steps of the resource data processing method provided in any one of the embodiments of the present application are implemented .
  • One or more computer-readable storage media storing computer-readable instructions.
  • the computer-readable storage media may be nonvolatile or volatile.
  • the computer-readable instructions are executed by one or more processors , Enabling one or more processors to implement the steps of the resource data processing method provided in any embodiment of the present application.
  • Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory.
  • Volatile memory may include random access memory (RAM) or external cache memory.
  • RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Channel (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
  • SRAM static RAM
  • DRAM dynamic RAM
  • SDRAM synchronous DRAM
  • DDRSDRAM double data rate SDRAM
  • ESDRAM enhanced SDRAM
  • SLDRAM synchronous chain Channel
  • memory bus Radbus direct RAM
  • RDRAM direct memory bus dynamic RAM
  • RDRAM memory bus dynamic RAM

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • General Business, Economics & Management (AREA)
  • Operations Research (AREA)
  • Educational Administration (AREA)
  • Game Theory and Decision Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种资源数据的处理方法,涉及人工智能技术领域,尤其涉及预测分析技术领域,包括:接收终端发送的查询请求;查询请求携带资源数据类型;获取与资源数据类型对应的预测指标,从数据库中获取与预测指标对应的待预测数据;根据目标数据标识符,从待预测数据中提取出对应的目标数据,作为与预测指标对应的目标数据;根据预设数据转化指令,对与预测指标对应的目标数据进行转化,得到与预测指标对应的参考值;将得到的与各个预测指标对应的参考值输入预先训练的资源数据预测模型,得到资源数据变化比例;根据资源数据变化比例生成资源数据预测信息,将资源数据预测信息发送至终端进行显示。

Description

资源数据的处理方法、装置、计算机设备和存储介质
本申请要求于2019年07月09日提交中国专利局,申请号为201910614250.7,申请名称为“资源数据的处理方法、装置、计算机设备和存储介质”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
本申请涉及预测分析技术领域,特别是涉及一种资源数据的处理方法、装置、计算机设备和存储介质。
背景技术
资源数据预测信息是指资源数据如利率在未来一段时间内的变化比例;不同资源数据类型对应的资源数据预测信息不一样;为了及时了解资源数据在未来一段时间的变化比例,对资源数据进行预测显得非常重要。
然而,目前对于资源数据变化比例的预测,一般是通过服务器对于人工收集的大量特定数据,比如经济数据、资金数据、政策数据等,进行单一的分析计算,初步预测资源数据在未来一段时间的变化趋势。然而,发明人意识到,资源数据变化趋势受多个因素影响,不同资源数据类型对应的主要影响因素不一样,导致人工收集的很多数据没有实际参考价值;且现有的服务器资源有限,若针对每一种资源数据类型,服务器都需要基于人工收集的大量特定数据进行计算评估,造成计算周期长,从而导致服务器的资源浪费,使得服务器的资源利用率下降。
发明内容
根据本申请公开的各种实施例,提供一种资源数据的处理方法、装置、计算机设备和存储介质。
一种资源数据的处理方法包括:
接收终端发送的查询请求;所述查询请求用于获取待预测资源数据的预测信息,所述查询请求携带所述待预测资源数据的资源数据类型;
获取与所述资源数据类型对应的预测指标,从数据库中获取与所述预测指标对应的待预测数据;
获取所述待预测数据的目标数据标识符,根据所述目标数据标识符,从所述待预测数据中提取出对应的目标数据,作为与所述预测指标对应的目标数据;
获取预设数据转化指令,根据所述预设数据转化指令,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值;
将得到的与所述预测指标对应的参考值输入预先训练的资源数据预测模型,得到与所述资源数据类型对应的资源数据变化比例;及
根据与所述资源数据类型对应的资源数据变化比例生成资源数据预测信息,将所述资源数据预测信息发送至所述终端进行显示。
一种资源数据的处理装置包括:
请求接收模块,用于接收终端发送的查询请求;所述查询请求用于获取待预测资源数据的预测信息,所述查询请求携带所述待预测资源数据的资源数据类型;
数据获取模块,用于获取与所述资源数据类型对应的预测指标,从数据库中获取与所述预测指标对应的待预测数据;
数据提取模块,用于获取所述待预测数据的目标数据标识符,根据所述目标数据标识符,从所述待预测数据中提取出对应的目标数据,作为与所述预测指标对应的目标数据;
数据转化模块,用于获取预设数据转化指令,根据所述预设数据转化指令,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值;
变化比例获取模块,用于将得到的与所述预测指标对应的参考值输入预先训练的资源数据预测模型,得到与所述资源数据类型对应的资源数据变化比例;及
信息生成模块,用于根据与所述资源数据类型对应的资源数据变化比例生成资源数据预测信息,将所述资源数据预测信息发送至所述终端进行显示。
一种计算机设备,包括存储器和一个或多个处理器,所述存储器中储存有计算机可读指令,所述计算机可读指令被所述处理器执行时,使得所述一个或多个处理器执行以下步骤:
接收终端发送的查询请求;所述查询请求用于获取待预测资源数据的预测信息,所述查询请求携带所述待预测资源数据的资源数据类型;
获取与所述资源数据类型对应的预测指标,从数据库中获取与所述预测指标对应的待预测数据;
获取所述待预测数据的目标数据标识符,根据所述目标数据标识符,从所述待预测数据中提取出对应的目标数据,作为与所述预测指标对应的目标数据;
获取预设数据转化指令,根据所述预设数据转化指令,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值;
将得到的与所述预测指标对应的参考值输入预先训练的资源数据预测模型,得到与所述资源数据类型对应的资源数据变化比例;及
根据与所述资源数据类型对应的资源数据变化比例生成资源数据预测信息,将所述资源数据预测信息发送至所述终端进行显示。
一个或多个存储有计算机可读指令的计算机可读存储介质,计算机可读指令被一个或多个处理器执行时,使得一个或多个处理器执行以下步骤:
接收终端发送的查询请求;所述查询请求用于获取待预测资源数据的预测信息,所述 查询请求携带所述待预测资源数据的资源数据类型;
获取与所述资源数据类型对应的预测指标,从数据库中获取与所述预测指标对应的待预测数据;
获取所述待预测数据的目标数据标识符,根据所述目标数据标识符,从所述待预测数据中提取出对应的目标数据,作为与所述预测指标对应的目标数据;
获取预设数据转化指令,根据所述预设数据转化指令,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值;
将得到的与所述预测指标对应的参考值输入预先训练的资源数据预测模型,得到与所述资源数据类型对应的资源数据变化比例;及
根据与所述资源数据类型对应的资源数据变化比例生成资源数据预测信息,将所述资源数据预测信息发送至所述终端进行显示。
上述资源数据的处理方法、装置、计算机设备和存储介质,服务器通过预先训练的资源数据预测模型,只对与资源数据类型的资源数据变化比例相关的预测指标对应的待预测数据进行评估分析,无需针对每一种资源数据类型,都对人工收集的大量特定数据进行计算评估,避免了服务器资源浪费,从而提高了服务器的资源利用率;同时,结合预先训练的资源数据预测模型,对资源数据类型相关的待预测数据进行分析,实现了对与资源数据类型对应的资源数据变化比例的全面跟踪评估的目的,进一步提高了得到的资源数据预测信息的准确性。
本申请的一个或多个实施例的细节在下面的附图和描述中提出。本申请的其它特征和优点将从说明书、附图以及权利要求书变得明显。
附图说明
为了更清楚地说明本申请实施例中的技术方案,下面将对实施例中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅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它的附图。
图1为根据一个或多个实施例中资源数据的处理方法的应用场景图;
图2为根据一个或多个实施例中资源数据的处理方法的流程示意图;
图3为根据一个或多个实施例中得到与资源数据类型对应的资源数据变化比例的步骤的流程示意图;
图4为根据一个或多个实施例中资源数据的处理装置的框图;
图5为根据一个或多个实施例中计算机设备的框图。
具体实施方式
为了使本申请的技术方案及优点更加清楚明白,以下结合附图及实施例,对本申请进行进一步详细说明。应当理解,此处描述的具体实施例仅仅用以解释本申请,并不用于限 定本申请。
本申请提供的资源数据的处理方法,可以应用于如图1所示的应用环境中。其中,终端110与服务器120通过网络进行通信。终端110安装了应用程序,用户通过该应用程序可以查询与资源数据类型对应的资源数据变化比例,比如与利率类型对应的利率波动比例。终端110响应用户对应用程序展示的查询界面的输入操作,生成用于获取资源数据预测信息的查询请求,并将该查询请求发送至服务器120。服务器120解析查询请求,得到用户输入的待预测资源数据的资源数据类型;获取与资源数据类型对应的预测指标,根据资源数据类型查询数据库,从数据库中获取与预测指标对应的待预测数据;获取待预测数据的目标数据标识符,根据目标数据标识符,从待预测数据中提取出对应的目标数据,作为与预测指标对应的目标数据;获取预设数据转化指令,根据预设数据转化指令,对与预测指标对应的目标数据进行转化,得到与预测指标对应的参考值;将得到的与各个预测指标对应的参考值输入预先训练的资源数据预测模型,得到与资源数据类型对应的资源数据变化比例;根据与资源数据类型对应的资源数据变化比例生成资源数据预测信息,将资源数据预测信息发送至终端110。终端110根据接收的资源数据预测信息,显示与用户输入的资源数据的资源数据类型对应的资源数据变化比例。其中,终端110可以但不限于是各种个人计算机、笔记本电脑、智能手机和平板电脑,服务器120可以用独立的服务器或者是多个服务器组成的服务器集群来实现。
在其中一个实施例中,如图2所示,提供了一种资源数据的处理方法,以该方法应用于图1中的服务器为例进行说明,包括以下步骤:
步骤S201,接收终端发送的查询请求;查询请求用于获取待预测资源数据的预测信息,查询请求携带待预测资源数据的资源数据类型。
在本步骤中,资源数据是指数值增益比率,表示数值增益的百分比,比如利率;资源数据预测信息是指待预测资源数据在未来一段时间内的变化比例,比如利率预测信息;资源数据类型是指用于标识待预测资源数据的信息,比如贷款利率、存款利率等。
终端安装了用于查询与资源数据类型对应的资源数据预测信息的应用程序,用户通过登录应用程序可以查询与资源数据类型对应的资源数据预测信息。终端响应用户对应用程序展示的查询界面的输入操作,生成用于获取资源数据预测信息的查询请求,并将该查询请求发送至服务器。服务器解析接收到的查询请求,得到待预测资源数据的资源数据类型。此外,用户还可以通过登录终端运行的浏览器,进入资源数据预测信息的查询界面,用户基于查询界面输入待预测资源数据的资源数据类型,通过终端将携带资源数据类型的查询请求发送至服务器。
步骤S202,获取与资源数据类型对应的预测指标,从数据库中获取与预测指标对应的待预测数据。
在本步骤中,预测指标是指对资源数据变化比例造成影响的指标,比如新闻、政策、货币供给、社会融资总量、专家研究报告等;不同资源数据类型对应的预测指标不一样。 服务器基于大数据,将预测指标作为搜索关键词,预先从互联网爬取与各个预测指标对应的数据,对爬取到的与各个预测指标对应的数据进行预处理操作,比如去除噪声、过滤干扰信息等,获取与各个预测指标对应的关键数据内容;将获取到的关键数据内容作为待预测数据,并按照预测指标分类存储至预先建立的数据库中;方便后续服务器通过查询数据库,获取与预测指标对应的待预测数据。
此外,在一定时间之后,比如2个月,服务器重新基于互联网获取与各个预测指标对应的待预测数据,对于同一预测指标,采用重新获取的待预测数据覆盖数据库中与该预测指标对应的原始数据,以及时更新数据库中的数据,有利于提高获取的与各个指标对应的待预测数据的准确性和时效性,进一步提高了后续得到的资源数据预测信息的准确性。
服务器从预设的已知资源数据类型中,筛选出与获取到的资源数据类型匹配的已知资源数据类型;获取与该已知资源数据类型对应的预测指标,将与该已知资源数据类型对应的预测指标作为与获取到的资源数据类型对应的预测指标;根据与获取到的资源数据类型对应的预测指标查询数据库,从数据库中获取与预测指标对应的待预测数据,方便后续服务器根据与预测指标对应的待预测数据确定与资源数据类型对应的资源数据变化比例,实现了对资源数据变化比例的全面跟踪评估,从而提高了得到的资源数据预测信息的准确性。
步骤S203,获取待预测数据的目标数据标识符,根据目标数据标识符,从待预测数据中提取出对应的目标数据,作为与预测指标对应的目标数据。
在本步骤中,目标数据标识符用于标识待预测数据中的关键信息,不同目标数据对应的目标数据标识符不一样;目标数据是指待预测数据中的关键信息;假设待预测数据是新闻数据,那么待预测数据中的目标数据是指新闻数据中的关键信息。
具体实现中,服务器预先获取待预测数据的目标数据标识符,从待预测数据中,提取出与目标数据标识符对应的目标数据,将该目标数据识别为与预测指标对应的目标数据。这样,有利于提取待预测数据中的目标数据,避免多余数据干扰,进一步减少了服务器需要处理的数据,从而提高了服务器的资源利用率。
步骤S204,获取预设数据转化指令,根据预设数据转化指令,对与预测指标对应的目标数据进行转化,得到与预测指标对应的参考值。
在本步骤中,数据转化指令是一种能够将与预测指标对应的待预测数据转化为对应的参考值的指令;不同预测指标,对应的数据转化指令不一样。参考值是指基于数据转化指令,由与预测指标对应的目标数据转化而成的数值;用于将目标数据转化为便于预测模型计算的数值。例如,将社会融资总量指标对应的目标数据转化为对应的社会融资总量级别,那么该社会融资总量级别即为社会融资总量指标对应的参考值。
具体实现中,服务器获取预设数据转化指令,根据预设数据转化指令,对与预测指标对应的目标数据进行转化,得到相应的数据级别,并将得到的数据级别识别为与预测指标对应的参考值;方便后续服务器根据得到的参考值,对资源数据变化比例进行预测,避免多余数据干扰,从而提高了得到的资源数据预测信息的准确性。
针对数字型目标数据,例如将社会融资总量指标对应的目标数据数据转化为对应的社会融资总量级别,并将得到的社会融资总量级别作为与社会融资总量指标对应的参考值;假设社会融资总量为c,则对应的社会融资总量级别为一级,即与社会融资总量指标对应的参考值为1;假设社会融资总量为d,则对应的社会融资总量级别为二级,即与社会融资总量指标对应的参考值为2。针对理论型目标数据,例如根据政策指标对应的目标数据对资源数据变化比例上涨的影响程度,将该目标数据转化为对应的政策级别,并将得到的政策级别作为与政策指标对应的参考值;比如政策指标对应的目标数据对资源数据变化比例上涨的影响程度较高,则对应的政策级别为一级,即与政策指标对应的参考值为1。
步骤S205,将得到的与预测指标对应的待预测数据输入预先训练的资源数据预测模型,得到与资源数据类型对应的资源数据变化比例。
在本步骤中,资源数据预测模型是一种能够根据输入的与预测指标对应的参考值,确定与资源数据类型对应的资源数据变化比例的模型。其中,资源数据变化比例是指数值增益比率在未来一段时间内的变化比例,比如利率变化比例,用于表示数值增益比率在未来一段时间内的波动趋势。
服务器将得到的与各个预测指标对应的参考值输入预先训练的资源数据预测模型中,通过资源数据预测模型对与各个预测指标对应的参考值进行计算分析,得到各个预测指标对应的资源数据预测结果;综合各个预测指标对应的资源数据预测结果,得到与资源数据类型对应的资源数据变化比例,从而提高了得到的资源数据预测信息的准确性。
步骤S206,根据与资源数据类型对应的资源数据变化比例生成资源数据预测信息,将资源数据预测信息发送至终端进行显示。
在本步骤中,服务器获取预设资源数据预测信息模板,将与资源数据类型对应的资源数据变化比例导入到预设资源数据预测信息模板中,生成对应的资源数据预测信息,并将生成的资源数据预测信息发送至对应的终端,通过终端的显示界面显示与资源数据类型对应的资源数据变化比例,方便用户直观、全面地了解资源数据类型的资源数据在未来一段时间内的变化比例,有利于用户对资源数据类型的资源数据变化比例进行全面跟踪评估。
上述资源数据的处理方法中,服务器通过预先训练的资源数据预测模型,只对与资源数据类型的资源数据变化比例相关的预测指标对应的待预测数据进行评估分析,无需针对每一种资源数据类型,都对人工收集的大量特定数据进行计算评估,避免了服务器资源浪费,从而提高了服务器的资源利用率;同时,结合预先训练的资源数据预测模型,对资源数据类型相关的待预测数据进行分析,实现了对与资源数据类型对应的资源数据变化比例的全面跟踪评估的目的,进一步提高了得到的资源数据预测信息的准确性。
在其中一个实施例中,上述步骤S202,获取与资源数据类型对应的预测指标的步骤具体包括:将资源数据类型与预设的已知资源数据类型进行匹配;若资源数据类型与已知资源数据类型匹配,获取与已知资源数据类型对应的预测指标;将获取的与已知资源数据类型对应的预测指标,作为与资源数据类型对应的预测指标。
本实施例中,服务器分别计算资源数据类型与预设的各个已知资源数据类型之间的匹配度;从预设的已知资源数据类型中,筛选出匹配度最大的已知资源数据类型,将匹配度最大的已知资源数据类型作为与获取到的资源数据类型成功匹配的已知资源数据类型;获取与匹配度最大的已知资源数据类型对应的预测指标,将与该已知资源数据类型对应的预测指标,作为与获取到的资源数据类型对应的预测指标;方便后续服务器通过查询数据库,获取与各个预测指标对应的待预测数据。
在其中一个实施例中,上述步骤S204,获取预设数据转化指令,根据预设数据转化指令,对与预测指标对应的目标数据进行转化,得到与预测指标对应的参考值,包括:提取预设数据转化指令中的数据转化规则;数据转化规则为目标数据与参考值之间的转化规则;根据数据转化规则,对与预测指标对应的目标数据进行转化,得到与预测指标对应的参考值。
本实施例中,数据转化规则是指将与预测指标对应的目标数据转化为相应的参考值的规则;不同预测指标,对应的数据转化规则不一样。
服务器获取数据转化规则的标识符,根据数据转化规则的标识符,从数据转化指令中提取出对应的数据转化规则;根据数据转化规则,对与预测指标对应的目标数据进行转化,得到相应的数据级别,并将得到的数据级别识别为与预测指标对应的参考值。通过本实施例,方便后续服务器根据得到的参考值,对资源数据变化比例进行预测,避免了多余数据干扰,从而提高了得到的资源数据预测信息的准确性;同时,减少了服务器需要处理的数据,有利于提高服务器的资源利用率。
进一步地,考虑到资源数据预测模型包括与各个预测指标对应的指标预测模型,那么服务器还可以通过与各个预测指标对应的指标预测模型,对对应的参考值进行分析,综合各个指标预测模型的资源数据变化子比例,以提高得到的资源数据预测信息的准确性。
在其中一个实施例中,如图3所示,将得到的与预测指标对应的参考值输入预先训练的资源数据预测模型,得到与资源数据类型对应的资源数据变化比例的步骤具体包括:
步骤S301,分别将与各个预测指标对应的参考值输入对应的指标预测模型,得到与各个预测指标对应的资源数据变化子比例。
步骤S302,获取预设的与各个指标预测模型对应的权重因子。
步骤S303,分别根据与各个指标预测模型对应的权重因子,对对应得到的资源数据变化子比例进行加权计算,得到与资源数据类型对应的资源数据变化比例。
本实施例中,指标预测模型是指能够根据输入的与预测指标对应的参考值,大致确定对应的资源数据变化比例的模型。比如,政策指标预测模型能够根据重大事件、重要政策等与政策相关的待预测数据对应的参考值,大致预测资源数据变化子比例;例如中美贸易战会影响人民币对美元汇率,汇率会影响利率在未来一段时间的变化趋势;需要说明的是,其他指标预测模型也是基于同样的原理实现,在此不再赘述。资源数据变化子比例是指根据与单一预测指标对应的待预测数据对应的参考值而大致预测出的资源数据变化比例,表 示在单一预测指标下,数值增益比率在未来一段时间内的变化比例。资源数据变化比例是指综合多个预测指标对应的资源数据变化子比例而预测出的变化总比例。权重因子是指各个指标预测模型输出的资源数据变化子比例对最终得到的资源数据变化比例的影响程度;权重因子越大,对最终得到的资源数据变化比例的影响程度越大。
例如,服务器分别将与各个预测指标对应的参考值输入对应的指标预测模型,通过指标预测模型对对应的参考值进行计算评估,确定对应的资源数据变化比例,并将确定的资源数据变化比例作为与预测指标对应的资源数据变化子比例;结合预设的与各个指标预测模型对应的权重因子,对对应得到的资源数据变化子比例进行加权计算,得到与资源数据类型对应的资源数据变化比例。假设预测指标为a,b,c,d,与预测指标a,b,c,d对应的资源数据变化子比例分别为y1,y2,y3,,y4,与预测指标a,b,c,d对应的指标预测模型的权重因子分别为m1,m2,m3,m4,则得到的与资源数据类型对应的资源数据变化比例为y=y1×m1+y2×m2+y3×m3+y4×m4。
在本实施例中,通过各个指标预测模型对各个预测指标对应的参考值进行分析,避免多余数据干扰,减少了服务器需要处理的数据,从而提高了服务器的资源利用率;同时,能够从多个角度对资源数据变化比例进行分析,从而实现了对资源数据变化比例进行全面跟踪评估的目的,避免出现误差,从而提高了得到的资源数据预测信息的准确性。
在其中一个实施例中,上述步骤S205,根据与资源数据类型对应的资源数据变化比例生成资源数据预测信息,将资源数据预测信息发送至终端进行显示,包括:获取各个预测指标在预设资源数据图谱模板中对应的导入位置;将与各个预测指标对应的待预测数据和资源数据变化子比例,导入到预设资源数据图谱模板中对应的导入位置,生成资源数据图谱;根据资源数据图谱以及与资源数据类型对应的资源数据变化比例生成资源数据参考信息;将资源数据图谱、与资源数据类型对应的资源数据变化比例和资源数据参考信息,依次导入到预设资源数据预测信息模板中,生成对应的资源数据预测信息,并将资源数据预测信息发送至终端进行显示。
本实施例中,资源数据参考信息是指根据生成的资源数据图谱以及与资源数据类型对应的资源数据变化比例总结而出的信息;例如哪些预测指标对资源数据变化比例影响较大,需要采取什么应对措施等;便于用户直接查看,无需通过人工总结。
具体实现中,服务器将生成的资源数据图谱、与资源数据类型对应的资源数据变化比例和资源数据参考信息,依次导入到预设资源数据预测信息模板中,生成对应的资源数据预测信息,并将资源数据预测信息发送至终端,通过终端的显示界面显示资源数据图谱、与资源数据类型对应的资源数据变化比例和资源数据参考信息;例如,通过终端的显示界面显示利率图谱、与利率类型对应的利率变化比例和利率参考信息。通过本实施例,可以方便用户直观、全面地了解资源数据在未来一段时间中的变化比例,有利于用户对与资源数据类型对应的资源数据变化比例进行全面跟踪评估。
进一步地,在根据与资源数据类型对应的资源数据变化比例生成资源数据预测信息, 将资源数据预测信息发送至终端进行显示之后,服务器还可以检测终端返回的资源数据预测信息的确认信息;若在预设时间范围内,检测到终端返回的资源数据预测信息的确认信息,则将资源数据预测信息标记为已发送;若在预设时间范围内,没有检测到终端返回的资源数据预测信息的确认信息,则重新将资源数据预测信息发送至对应的终端。通过在预设时间范围内,服务器检测终端是否返回资源数据预测信息的确认信息,可以有效地判断终端是否接收到资源数据预测信息,避免重复发送资源数据预测信息而浪费服务器资源,进一步提高了服务器的资源利用率。
此外,为了进一步提高指标预测模型的资源数据预测准确率,可以对指标预测模型进行多次训练。在其中一个实施例中,指标预测模型通过下述方法得到:分别获取与各个待训练的指标预测模型对应的样本预测数据;根据样本预测数据对对应的待训练的指标预测模型进行训练,得到训练后的指标预测模型;获取训练后的指标预测模型输出的资源数据变化子比例与对应的实际资源数据变化子比例之间的预测误差;当预测误差大于或等于对应的预设阈值时,根据预测误差对待训练的指标预测模型进行反复训练,直到根据训练后的指标预测模型得到的预测误差小于对应的预设阈值。
比如,当预测误差大于或等于对应的预设阈值时,服务器根据预测误差调整指标预测模型的参数;根据调整后的指标预测模型对样本预测数据进行再次评估,获取根据调整后的指标预测模型得到的资源数据变化子比例与对应的实际资源数据变化子比例之间的预测误差,根据预测误差对指标预测模型的参数进行再次调整,以对指标预测模型进行再次训练,直到根据训练后的指标预测模型得到的资源数据变化子比例与对应的实际资源数据变化子比例之间的预测误差小于对应的预设阈值;当通过各个指标预测模型得到的预测误差均小于对应的预设阈值时,分别获取当前的各个指标预测模型;将当前的各个指标预测模型,作为各个训练好的指标预测模型。本实施例中,服务器根据预测误差,分别对各个指标预测模型进行多次训练,有利于通过各个指标预测模型输出更准确的资源数据变化子比例,从而提高了各个指标预测模型的资源数据预测准确率,有利于提高后续得到的资源数据预测信息的准确性。
进一步地,服务器还可以对资源数据预测模型进行多次训练,以提高得到的资源数据预测信息的准确性。在其中一个实施例中,资源数据预测模型通过下述方法得到:分别获取与各个训练好的指标预测模型对应的样本预测数据和权重因子;根据各个训练好的指标预测模型以及对应的样本预测数据和权重因子,对待训练的资源数据预测模型进行训练,得到训练后的资源数据预测模型;获取训练后的资源数据预测模型输出的资源数据变化比例与对应的实际资源数据变化比例之间的第一预测误差;当第一预测误差大于或等于第一预设阈值时,根据第一预测误差调整与各个训练好的指标预测模型对应的权重因子,并根据调整后的权重因子,对待训练的资源数据预测模型进行反复训练,直到根据训练后的资源数据预测模型得到的第一预测误差小于第一预设阈值。
比如,当第一预测误差大于或等于第一预设阈值时,服务器根据第一预测误差调整与 各个训练好的指标预测模型对应的权重因子,并根据调整后的权重因子,对待训练的资源数据预测模型进行再次训练;获取根据再次训练后的资源数据预测模型得到的资源数据变化比例与对应的实际资源数据变化比例之间的第一预测误差,根据第一预测误差对各个训练好的指标预测模型对应的权重因子进行再次调整,以对待训练的资源数据预测模型进行再次训练,直到根据训练后的资源数据预测模型得到的资源数据变化比例与对应的实际资源数据变化比例之间的第一预测误差小于第一预设阈值;当得到的第一预测误差小于第一预设阈值时,获取当前的资源数据预测模型,作为训练好的资源数据预测模型;并将当前的各个权重因子,分别作为预设的与各个训练好的指标预测模型对应的权重因子。本实施例中,通过不断调整各个训练好的指标预测模型的权重因子,以对资源数据预测模型进行多次训练,有利于通过资源数据预测模型输出更准确的资源数据变化比例,从而提高了得到的资源数据预测信息的准确性。
应该理解的是,虽然图2-3的流程图中的各个步骤按照箭头的指示依次显示,但是这些步骤并不是必然按照箭头指示的顺序依次执行。除非本文中有明确的说明,这些步骤的执行并没有严格的顺序限制,这些步骤可以以其它的顺序执行。而且,图2-3中的至少一部分步骤可以包括多个子步骤或者多个阶段,这些子步骤或者阶段并不必然是在同一时刻执行完成,而是可以在不同的时刻执行,这些子步骤或者阶段的执行顺序也不必然是依次进行,而是可以与其它步骤或者其它步骤的子步骤或者阶段的至少一部分轮流或者交替地执行。
在其中一个实施例中,如图4所示,提供了一种资源数据的处理装置,包括:请求接收模块410、数据获取模块420、数据提取模块430、数据转化模块440、变化比例获取模块450和信息生成模块460,其中:
请求接收模块410,用于接收终端发送的查询请求;查询请求用于获取待预测资源数据的预测信息,查询请求携带待预测资源数据的资源数据类型。
数据获取模块420,用于获取与资源数据类型对应的预测指标,从数据库中获取与预测指标对应的待预测数据。
数据提取模块430,用于获取待预测数据的目标数据标识符,根据目标数据标识符,从待预测数据中提取出对应的目标数据,作为与预测指标对应的目标数据。
数据转化模块440,用于获取预设数据转化指令,根据预设数据转化指令,对与预测指标对应的目标数据进行转化,得到与预测指标对应的参考值。
变化比例获取模块450,用于将获取的与各个预测指标对应的参考值输入预先训练的资源数据预测模型,得到与资源数据类型对应的资源数据变化比例。
信息生成模块460,用于根据与资源数据类型对应的资源数据变化比例生成资源数据预测信息,将资源数据预测信息发送至终端进行显示。
在其中一个实施例中,数据获取模块还用于将资源数据类型与预设的已知资源数据类型进行匹配;若资源数据类型与已知资源数据类型匹配,获取与已知资源数据类型对应的 预测指标;将获取的与已知资源数据类型对应的预测指标,作为与资源数据类型对应的预测指标。
在其中一个实施例中,数据转化模块还用于提取预设数据转化指令中的数据转化规则;数据转化规则为目标数据与参考值之间的转化规则;根据数据转化规则,对与预测指标对应的目标数据进行转化,得到与预测指标对应的参考值。
在其中一个实施例中,变化比例获取模块还用于分别将与各个预测指标对应的参考值输入对应的指标预测模型,得到与各个预测指标对应的资源数据变化子比例;获取预设的与各个指标预测模型对应的权重因子;分别根据与各个指标预测模型对应的权重因子,对对应得到的资源数据变化子比例进行加权计算,得到与资源数据类型对应的资源数据变化比例。
在其中一个实施例中,信息生成模块还用于获取各个预测指标在预设资源数据图谱模板中对应的导入位置;将与各个预测指标对应的待预测数据和资源数据变化子比例,导入到预设资源数据图谱模板中对应的导入位置,生成资源数据图谱;根据资源数据图谱以及与资源数据类型对应的资源数据变化比例生成资源数据参考信息;将资源数据图谱、与资源数据类型对应的资源数据变化比例和资源数据参考信息,依次导入到预设资源数据预测信息模板中,生成对应的资源数据预测信息,并将资源数据预测信息发送至终端进行显示。
在其中一个实施例中,资源数据的处理装置还包括指标预测模型训练模块,用于分别获取与各个待训练的指标预测模型对应的样本预测数据;根据样本预测数据对对应的待训练的指标预测模型进行训练,得到训练后的指标预测模型;获取训练后的指标预测模型输出的资源数据变化子比例与对应的实际资源数据变化子比例之间的预测误差;当预测误差大于或等于对应的预设阈值时,根据预测误差对待训练的指标预测模型进行反复训练,直到根据训练后的指标预测模型得到的预测误差小于对应的预设阈值。
在其中一个实施例中,资源数据的处理装置还包括资源数据模型训练模块,用于分别获取与各个训练好的指标预测模型对应的样本预测数据和权重因子;根据各个训练好的指标预测模型以及对应的样本预测数据和权重因子,对待训练的资源数据预测模型进行训练,得到训练后的资源数据预测模型;获取训练后的资源数据预测模型输出的资源数据变化比例与对应的实际资源数据变化比例之间的第一预测误差;当第一预测误差大于或等于第一预设阈值时,根据第一预测误差调整与各个训练好的指标预测模型对应的权重因子,并根据调整后的权重因子,对待训练的资源数据预测模型进行反复训练,直到根据训练后的资源数据预测模型得到的第一预测误差小于第一预设阈值。
上述各个实施例,资源数据的处理装置通过预先训练的资源数据预测模型,只对与资源数据类型的资源数据变化比例相关的预测指标对应的待预测数据进行评估分析,无需针对每一种资源数据类型,都对人工收集的大量特定数据进行计算评估,避免了服务器资源浪费,从而提高了服务器的资源利用率;同时,结合预先训练的资源数据预测模型,对资源数据类型相关的待预测数据进行分析,实现了对与资源数据类型对应的资源数据变化比 例的全面跟踪评估的目的,进一步提高了得到的资源数据预测信息的准确性。
关于资源数据的处理装置的具体限定可以参见上文中对于资源数据的处理方法的限定,在此不再赘述。上述资源数据的处理装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中,也可以以软件形式存储于计算机设备中的存储器中,以便于处理器调用执行以上各个模块对应的操作。
在其中一个实施例中,提供了一种计算机设备,该计算机设备可以是服务器,其内部结构图可以如图5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、网络接口和数据库。其中,该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性或易失性存储介质、内存储器。该非易失性或易失性存储介质存储有操作系统、计算机可读指令和数据库。该内存储器为非易失性或易失性存储介质中的操作系统和计算机可读指令的运行提供环境。该计算机设备的数据库用于存储与各个评估指标对应的待评估数据。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机可读指令被处理器执行时以实现一种资源数据的处理方法。
本领域技术人员可以理解,图5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。
一种计算机设备,包括存储器和一个或多个处理器,存储器中储存有计算机可读指令,计算机可读指令被处理器执行时实现本申请任意一个实施例中提供的资源数据的处理方法的步骤。
一个或多个存储有计算机可读指令的计算机可读存储介质,所述计算机可读存储介质可以是非易失性,也可以是易失性,计算机可读指令被一个或多个处理器执行时,使得一个或多个处理器实现本申请任意一个实施例中提供的资源数据的处理方法的步骤。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机可读指令来指令相关的硬件来完成,所述的计算机可读指令可存储于一计算机可读取存储介质中,该计算机可读指令在执行时,可包括如上述各方法的实施例的流程。其中,本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用,均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限,RAM以多种形式可得,诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchlink)DRAM(SLDRAM)、存储器总线(Rambus)直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。
以上实施例的各技术特征可以进行任意的组合,为使描述简洁,未对上述实施例中的 各个技术特征所有可能的组合都进行描述,然而,只要这些技术特征的组合不存在矛盾,都应当认为是本说明书记载的范围。
以上所述实施例仅表达了本申请的几种实施方式,其描述较为具体和详细,但并不能因此而理解为对发明专利范围的限制。应当指出的是,对于本领域的普通技术人员来说,在不脱离本申请构思的前提下,还可以做出若干变形和改进,这些都属于本申请的保护范围。因此,本申请专利的保护范围应以所附权利要求为准。

Claims (20)

  1. 一种资源数据的处理方法,包括:
    接收终端发送的查询请求;所述查询请求用于获取待预测资源数据的预测信息,所述查询请求携带所述待预测资源数据的资源数据类型;
    获取与所述资源数据类型对应的预测指标,从数据库中获取与所述预测指标对应的待预测数据;
    获取所述待预测数据的目标数据标识符,根据所述目标数据标识符,从所述待预测数据中提取出对应的目标数据,作为与所述预测指标对应的目标数据;
    获取预设数据转化指令,根据所述预设数据转化指令,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值;
    将得到的与所述预测指标对应的参考值输入预先训练的资源数据预测模型,得到与所述资源数据类型对应的资源数据变化比例;及
    根据与所述资源数据类型对应的资源数据变化比例生成资源数据预测信息,将所述资源数据预测信息发送至所述终端进行显示。
  2. 根据权利要求1所述的方法,其中,所述获取与所述资源数据类型对应的预测指标,包括:
    将所述资源数据类型与预设的已知资源数据类型进行匹配;
    若所述资源数据类型与所述已知资源数据类型匹配,获取与所述已知资源数据类型对应的预测指标;及
    将获取的与所述已知资源数据类型对应的预测指标,作为与所述资源数据类型对应的预测指标。
  3. 根据权利要求1所述的方法,其中,所述获取预设数据转化指令,根据所述预设数据转化指令,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值,包括:
    提取所述预设数据转化指令中的数据转化规则;所述数据转化规则为目标数据与参考值之间的转化规则;及
    根据所述数据转化规则,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值。
  4. 根据权利要求1所述的方法,其中,所述资源数据预测模型包括与各个所述预测指标对应的指标预测模型;
    所述将得到的与所述预测指标对应的参考值输入预先训练的资源数据预测模型,得到与所述资源数据类型对应的资源数据变化比例,包括:
    分别将与各个所述预测指标对应的参考值输入对应的指标预测模型,得到与各个所述预测指标对应的资源数据变化子比例;
    获取预设的与各个所述指标预测模型对应的权重因子;及
    分别根据与各个所述指标预测模型对应的权重因子,对对应得到的资源数据变化子比例进行加权计算,得到与所述资源数据类型对应的资源数据变化比例。
  5. 根据权利要求1所述的方法,其中,所述根据与所述资源数据类型对应的资源数据变化比例生成资源数据预测信息,将所述资源数据预测信息发送至所述终端进行显示,包括:
    获取各个所述预测指标在预设资源数据图谱模板中对应的导入位置;
    将与各个所述预测指标对应的待预测数据和资源数据变化子比例,导入到所述预设资源数据图谱模板中对应的导入位置,生成资源数据图谱;
    根据所述资源数据图谱以及与所述资源数据类型对应的资源数据变化比例生成资源数据参考信息;及
    将所述资源数据图谱、与所述资源数据类型对应的资源数据变化比例和所述资源数据参考信息,依次导入到预设资源数据预测信息模板中,生成对应的资源数据预测信息,并将所述资源数据预测信息发送至所述终端进行显示。
  6. 根据权利要求1至5任意一项所述的方法,其中,所述指标预测模型通过下述方法得到:
    分别获取与各个待训练的指标预测模型对应的样本预测数据;
    根据样本预测数据对对应的待训练的指标预测模型进行训练,得到训练后的指标预测模型;
    获取所述训练后的指标预测模型输出的资源数据变化子比例与对应的实际资源数据变化子比例之间的预测误差;及
    当所述预测误差大于或等于对应的预设阈值时,根据所述预测误差对所述待训练的指标预测模型进行反复训练,直到根据训练后的指标预测模型得到的预测误差小于对应的预设阈值。
  7. 根据权利要求6所述的方法,其中,所述资源数据预测模型通过下述方法得到:
    分别获取与各个所述训练好的指标预测模型对应的样本预测数据和权重因子;
    根据各个所述训练好的指标预测模型以及对应的样本预测数据和权重因子,对待训练的资源数据预测模型进行训练,得到训练后的资源数据预测模型;
    获取所述训练后的资源数据预测模型输出的资源数据变化比例与对应的实际资源数据变化比例之间的第一预测误差;及
    当所述第一预测误差大于或等于第一预设阈值时,根据所述第一预测误差调整与各个所述训练好的指标预测模型对应的权重因子,并根据调整后的权重因子,对所述待训练的资源数据预测模型进行反复训练,直到根据训练后的资源数据预测模型得到的第一预测误差小于所述第一预设阈值。
  8. 一种资源数据的处理装置,包括:
    请求接收模块,用于接收终端发送的查询请求;所述查询请求用于获取待预测资源数 据的预测信息,所述查询请求携带所述待预测资源数据的资源数据类型;
    数据获取模块,用于获取与所述资源数据类型对应的预测指标,从数据库中获取与所述预测指标对应的待预测数据;
    数据提取模块,用于获取所述待预测数据的目标数据标识符,根据所述目标数据标识符,从所述待预测数据中提取出对应的目标数据,作为与所述预测指标对应的目标数据;
    数据转化模块,用于获取预设数据转化指令,根据所述预设数据转化指令,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值;
    变化比例获取模块,用于将得到的与所述预测指标对应的参考值输入预先训练的资源数据预测模型,得到与所述资源数据类型对应的资源数据变化比例;及
    信息生成模块,用于根据与所述资源数据类型对应的资源数据变化比例生成资源数据预测信息,将所述资源数据预测信息发送至所述终端进行显示。
  9. 一种计算机设备,包括存储器及一个或多个处理器,所述存储器中储存有计算机可读指令,所述计算机可读指令被所述一个或多个处理器执行时,使得所述一个或多个处理器执行以下步骤:
    接收终端发送的查询请求;所述查询请求用于获取待预测资源数据的预测信息,所述查询请求携带所述待预测资源数据的资源数据类型;
    获取与所述资源数据类型对应的预测指标,从数据库中获取与所述预测指标对应的待预测数据;
    获取所述待预测数据的目标数据标识符,根据所述目标数据标识符,从所述待预测数据中提取出对应的目标数据,作为与所述预测指标对应的目标数据;
    获取预设数据转化指令,根据所述预设数据转化指令,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值;
    将得到的与所述预测指标对应的参考值输入预先训练的资源数据预测模型,得到与所述资源数据类型对应的资源数据变化比例;及
    根据与所述资源数据类型对应的资源数据变化比例生成资源数据预测信息,将所述资源数据预测信息发送至所述终端进行显示。
  10. 根据权利要求9所述的计算机设备,其中,所述处理器执行所述计算机可读指令时还执行以下步骤:
    将所述资源数据类型与预设的已知资源数据类型进行匹配;
    若所述资源数据类型与所述已知资源数据类型匹配,获取与所述已知资源数据类型对应的预测指标;及
    将获取的与所述已知资源数据类型对应的预测指标,作为与所述资源数据类型对应的预测指标。
  11. 根据权利要求9所述的计算机设备,其中,所述处理器执行所述计算机可读指令时还执行以下步骤:
    提取所述预设数据转化指令中的数据转化规则;所述数据转化规则为目标数据与参考值之间的转化规则;及
    根据所述数据转化规则,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值。
  12. 根据权利要求9所述的计算机设备,其中,所述处理器执行所述计算机可读指令时还执行以下步骤:
    获取各个所述预测指标在预设资源数据图谱模板中对应的导入位置;
    将与各个所述预测指标对应的待预测数据和资源数据变化子比例,导入到所述预设资源数据图谱模板中对应的导入位置,生成资源数据图谱;
    根据所述资源数据图谱以及与所述资源数据类型对应的资源数据变化比例生成资源数据参考信息;及
    将所述资源数据图谱、与所述资源数据类型对应的资源数据变化比例和所述资源数据参考信息,依次导入到预设资源数据预测信息模板中,生成对应的资源数据预测信息,并将所述资源数据预测信息发送至所述终端进行显示。
  13. 根据权利要求9至12任一项所述的计算机设备,其中,所述处理器执行所述计算机可读指令时还执行以下步骤:
    分别获取与各个待训练的指标预测模型对应的样本预测数据;
    根据样本预测数据对对应的待训练的指标预测模型进行训练,得到训练后的指标预测模型;
    获取所述训练后的指标预测模型输出的资源数据变化子比例与对应的实际资源数据变化子比例之间的预测误差;及
    当所述预测误差大于或等于对应的预设阈值时,根据所述预测误差对所述待训练的指标预测模型进行反复训练,直到根据训练后的指标预测模型得到的预测误差小于对应的预设阈值。
  14. 根据权利要求13所述的计算机设备,其中,所述处理器执行所述计算机可读指令时还执行以下步骤:
    分别获取与各个所述训练好的指标预测模型对应的样本预测数据和权重因子;
    根据各个所述训练好的指标预测模型以及对应的样本预测数据和权重因子,对待训练的资源数据预测模型进行训练,得到训练后的资源数据预测模型;
    获取所述训练后的资源数据预测模型输出的资源数据变化比例与对应的实际资源数据变化比例之间的第一预测误差;及
    当所述第一预测误差大于或等于第一预设阈值时,根据所述第一预测误差调整与各个所述训练好的指标预测模型对应的权重因子,并根据调整后的权重因子,对所述待训练的资源数据预测模型进行反复训练,直到根据训练后的资源数据预测模型得到的第一预测误差小于所述第一预设阈值。
  15. 一个或多个存储有计算机可读指令的计算机可读存储介质,所述计算机可读指令被一个或多个处理器执行时,使得所述一个或多个处理器执行以下步骤:
    接收终端发送的查询请求;所述查询请求用于获取待预测资源数据的预测信息,所述查询请求携带所述待预测资源数据的资源数据类型;
    获取与所述资源数据类型对应的预测指标,从数据库中获取与所述预测指标对应的待预测数据;
    获取所述待预测数据的目标数据标识符,根据所述目标数据标识符,从所述待预测数据中提取出对应的目标数据,作为与所述预测指标对应的目标数据;
    获取预设数据转化指令,根据所述预设数据转化指令,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值;
    将得到的与所述预测指标对应的参考值输入预先训练的资源数据预测模型,得到与所述资源数据类型对应的资源数据变化比例;及
    根据与所述资源数据类型对应的资源数据变化比例生成资源数据预测信息,将所述资源数据预测信息发送至所述终端进行显示。
  16. 根据权利要求15所述的存储介质,其中,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    将所述资源数据类型与预设的已知资源数据类型进行匹配;
    若所述资源数据类型与所述已知资源数据类型匹配,获取与所述已知资源数据类型对应的预测指标;及
    将获取的与所述已知资源数据类型对应的预测指标,作为与所述资源数据类型对应的预测指标。
  17. 根据权利要求15所述的存储介质,其中,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    提取所述预设数据转化指令中的数据转化规则;所述数据转化规则为目标数据与参考值之间的转化规则;及
    根据所述数据转化规则,对与所述预测指标对应的目标数据进行转化,得到与所述预测指标对应的参考值。
  18. 根据权利要求15所述的存储介质,其中,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    获取各个所述预测指标在预设资源数据图谱模板中对应的导入位置;
    将与各个所述预测指标对应的待预测数据和资源数据变化子比例,导入到所述预设资源数据图谱模板中对应的导入位置,生成资源数据图谱;
    根据所述资源数据图谱以及与所述资源数据类型对应的资源数据变化比例生成资源数据参考信息;及
    将所述资源数据图谱、与所述资源数据类型对应的资源数据变化比例和所述资源数据 参考信息,依次导入到预设资源数据预测信息模板中,生成对应的资源数据预测信息,并将所述资源数据预测信息发送至所述终端进行显示。
  19. 根据权利要求15至18任一项所述的存储介质,其中,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    分别获取与各个待训练的指标预测模型对应的样本预测数据;
    根据样本预测数据对对应的待训练的指标预测模型进行训练,得到训练后的指标预测模型;
    获取所述训练后的指标预测模型输出的资源数据变化子比例与对应的实际资源数据变化子比例之间的预测误差;及
    当所述预测误差大于或等于对应的预设阈值时,根据所述预测误差对所述待训练的指标预测模型进行反复训练,直到根据训练后的指标预测模型得到的预测误差小于对应的预设阈值。
  20. 根据权利要求19所述的存储介质,其中,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    分别获取与各个所述训练好的指标预测模型对应的样本预测数据和权重因子;
    根据各个所述训练好的指标预测模型以及对应的样本预测数据和权重因子,对待训练的资源数据预测模型进行训练,得到训练后的资源数据预测模型;
    获取所述训练后的资源数据预测模型输出的资源数据变化比例与对应的实际资源数据变化比例之间的第一预测误差;及
    当所述第一预测误差大于或等于第一预设阈值时,根据所述第一预测误差调整与各个所述训练好的指标预测模型对应的权重因子,并根据调整后的权重因子,对所述待训练的资源数据预测模型进行反复训练,直到根据训练后的资源数据预测模型得到的第一预测误差小于所述第一预设阈值。
PCT/CN2020/098879 2019-07-09 2020-06-29 资源数据的处理方法、装置、计算机设备和存储介质 WO2021004324A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910614250.7A CN110489630B (zh) 2019-07-09 2019-07-09 资源数据的处理方法、装置、计算机设备和存储介质
CN201910614250.7 2019-07-09

Publications (1)

Publication Number Publication Date
WO2021004324A1 true WO2021004324A1 (zh) 2021-01-14

Family

ID=68546748

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/098879 WO2021004324A1 (zh) 2019-07-09 2020-06-29 资源数据的处理方法、装置、计算机设备和存储介质

Country Status (2)

Country Link
CN (1) CN110489630B (zh)
WO (1) WO2021004324A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113449794A (zh) * 2021-06-26 2021-09-28 平安科技(深圳)有限公司 基于智能决策的政策预测方法、装置、设备及存储介质
CN113628753A (zh) * 2021-08-09 2021-11-09 海南大学 面向意图计算与推理的dikw资源分析方法及系统
CN113904940A (zh) * 2021-09-03 2022-01-07 深圳市雷鸟网络传媒有限公司 资源调整方法、装置、电子设备和计算机可读存储介质
CN114840348A (zh) * 2022-07-01 2022-08-02 石家庄学院 一种用于计算机的资源等级确定方法及系统

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110489630B (zh) * 2019-07-09 2024-08-20 平安科技(深圳)有限公司 资源数据的处理方法、装置、计算机设备和存储介质
CN111414699A (zh) * 2020-03-25 2020-07-14 平安资产管理有限责任公司 信息分析预测方法、装置、计算机系统及可读存储介质
CN112632381B (zh) * 2020-12-24 2021-10-01 中电金信软件有限公司 信息查询方法、装置、计算机设备和存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103337034A (zh) * 2013-06-08 2013-10-02 安徽磬鑫软件科技有限公司 基于蒙特卡洛-灰色预测混合算法的股票价格预测方法
CN104834755A (zh) * 2015-06-01 2015-08-12 山大地纬软件股份有限公司 一种医疗保险基金的预测数据监控系统及其方法
US20170213280A1 (en) * 2016-01-27 2017-07-27 Huawei Technologies Co., Ltd. System and method for prediction using synthetic features and gradient boosted decision tree
CN107818386A (zh) * 2017-11-28 2018-03-20 国网湖南省电力有限公司 电网企业经营利润预测方法
CN108011740A (zh) * 2016-10-28 2018-05-08 腾讯科技(深圳)有限公司 一种媒体流量数据处理方法和装置
CN110489630A (zh) * 2019-07-09 2019-11-22 平安科技(深圳)有限公司 资源数据的处理方法、装置、计算机设备和存储介质

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104899405B (zh) * 2014-03-04 2019-03-26 上海携程商务有限公司 数据预测方法和系统以及报警方法和系统
CN109032914A (zh) * 2018-09-06 2018-12-18 掌阅科技股份有限公司 资源占用数据预测方法、电子设备、存储介质
CN109523068A (zh) * 2018-10-31 2019-03-26 平安直通咨询有限公司 房屋价值预测方法、装置、计算机设备和存储介质
CN109935338A (zh) * 2019-03-07 2019-06-25 平安科技(深圳)有限公司 基于机器学习的数据预测处理方法、装置和计算机设备

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103337034A (zh) * 2013-06-08 2013-10-02 安徽磬鑫软件科技有限公司 基于蒙特卡洛-灰色预测混合算法的股票价格预测方法
CN104834755A (zh) * 2015-06-01 2015-08-12 山大地纬软件股份有限公司 一种医疗保险基金的预测数据监控系统及其方法
US20170213280A1 (en) * 2016-01-27 2017-07-27 Huawei Technologies Co., Ltd. System and method for prediction using synthetic features and gradient boosted decision tree
CN108011740A (zh) * 2016-10-28 2018-05-08 腾讯科技(深圳)有限公司 一种媒体流量数据处理方法和装置
CN107818386A (zh) * 2017-11-28 2018-03-20 国网湖南省电力有限公司 电网企业经营利润预测方法
CN110489630A (zh) * 2019-07-09 2019-11-22 平安科技(深圳)有限公司 资源数据的处理方法、装置、计算机设备和存储介质

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113449794A (zh) * 2021-06-26 2021-09-28 平安科技(深圳)有限公司 基于智能决策的政策预测方法、装置、设备及存储介质
CN113628753A (zh) * 2021-08-09 2021-11-09 海南大学 面向意图计算与推理的dikw资源分析方法及系统
CN113628753B (zh) * 2021-08-09 2023-04-14 海南大学 面向意图计算与推理的dikw资源分析方法及系统
CN113904940A (zh) * 2021-09-03 2022-01-07 深圳市雷鸟网络传媒有限公司 资源调整方法、装置、电子设备和计算机可读存储介质
CN114840348A (zh) * 2022-07-01 2022-08-02 石家庄学院 一种用于计算机的资源等级确定方法及系统

Also Published As

Publication number Publication date
CN110489630A (zh) 2019-11-22
CN110489630B (zh) 2024-08-20

Similar Documents

Publication Publication Date Title
WO2021004324A1 (zh) 资源数据的处理方法、装置、计算机设备和存储介质
CN109816221B (zh) 项目风险决策方法、装置、计算机设备和存储介质
WO2021184554A1 (zh) 数据库异常监测方法、装置、计算机装置及存储介质
WO2021004318A1 (zh) 资源数据的处理方法、装置、计算机设备和存储介质
Azzeh A replicated assessment and comparison of adaptation techniques for analogy-based effort estimation
CN113011973B (zh) 基于智能合约数据湖的金融交易监管模型的方法及设备
CN117350399A (zh) 用于管理机器学习模型的计算机实现的方法和计算设备
WO2021098384A1 (zh) 一种数据异常检测方法及装置
CN110489622B (zh) 对象信息的分享方法、装置、计算机设备和存储介质
CN109636212B (zh) 作业实际运行时间的预测方法
CN113298121B (zh) 基于多数据源建模的消息发送方法、装置和电子设备
US20190220924A1 (en) Method and device for determining key variable in model
CN110264270A (zh) 一种行为预测方法、装置、设备和存储介质
CN116561542B (zh) 模型的优化训练系统、方法以及相关装置
CN112668822A (zh) 科技成果转化平台共享系统、方法、存储介质、手机app
WO2021012861A1 (zh) 数据查询耗时评估方法、装置、计算机设备和存储介质
CN117764617A (zh) 一种基于大数据分析的物业运行成本预测方法和装置
Yin et al. An asymptotic statistical learning algorithm for prediction of key trading events
CN117540336A (zh) 时间序列预测方法、装置及电子设备
CN114610590A (zh) 作业运行时长确定方法、装置、设备及存储介质
Zhang et al. A combinational QoS-prediction approach based on RBF neural network
WO2021097494A2 (en) Distributed training of multi-modal machine learning models
CN118196567B (zh) 基于大语言模型的数据评价方法、装置、设备及存储介质
CN117556264B (zh) 一种评估模型的训练方法、装置及电子设备
CN112801372B (zh) 数据处理方法、装置、电子设备和可读存储介质

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: 20837412

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 20837412

Country of ref document: EP

Kind code of ref document: A1