CN117196827A - Method, device, equipment and storage medium for predicting early compensation rate of house loan - Google Patents
Method, device, equipment and storage medium for predicting early compensation rate of house loan Download PDFInfo
- Publication number
- CN117196827A CN117196827A CN202311208866.7A CN202311208866A CN117196827A CN 117196827 A CN117196827 A CN 117196827A CN 202311208866 A CN202311208866 A CN 202311208866A CN 117196827 A CN117196827 A CN 117196827A
- Authority
- CN
- China
- Prior art keywords
- historical data
- compensation rate
- early
- processed
- loan
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000012216 screening Methods 0.000 claims description 22
- 238000004088 simulation Methods 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000002372 labelling Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 230000002028 premature Effects 0.000 description 6
- 238000005070 sampling Methods 0.000 description 6
- 230000000875 corresponding effect Effects 0.000 description 5
- 238000004138 cluster model Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000000737 periodic effect Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Landscapes
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention discloses a house loan early compensation rate prediction method, a device, equipment and a storage medium, wherein the method comprises the following steps: clustering the historical data set to be processed according to a preset clustering model, and extracting target historical data from the clustered historical data set to be processed; and forming a simulated asset pool according to the target historical data, and predicting the house loan early compensation rate of the simulated asset pool. According to the invention, the historical data sets to be processed are clustered according to the preset clustering model, the data of the same type are divided into the same type, the historical data sets to be processed are sampled according to the divided types to obtain the target historical data sets, a simulated asset pool is formed according to the target historical data sets, and the early compensation rate of the house and the loan is predicted based on the simulated asset pool, so that the early compensation rate is not required to be fixed in a certain interval according to expert experience, and can not be timely adjusted according to market trend, and the prediction accuracy of the early compensation rate of the house and the loan is improved.
Description
Technical Field
The invention relates to the technical field of finance, in particular to a method, a device, equipment and a storage medium for predicting early compensation rate of a house credit.
Background
Due to the low risk characteristics of the individual housing mortgage, early compensation rate becomes the most dominant factor in the securities deadline and cash flow of the housing mortgage securities (Residential Mortgage-Backed Securities, RMBS), and whether early compensation rate prediction is accurate determines the safety of product redemption and the accuracy of expected benefits directly affects RMBS product market acceptance.
Professional investors, intermediaries, etc. in the market are limited by technical reserves and data availability, simply fix the predicted value of early compensation rate to 8% or 10% according to expert experience, and there are obvious limitations in fixing early compensation rate.
The foregoing is provided merely for the purpose of facilitating understanding of the technical solutions of the present invention and is not intended to represent an admission that the foregoing is prior art.
Disclosure of Invention
The invention mainly aims to provide a method, a device, equipment and a storage medium for predicting early compensation rate of a house and aims to solve the technical problem of how to improve the accuracy of the early compensation rate of the house and the credit.
In order to achieve the above object, the present invention provides a method for predicting a premature rate of a house, the method comprising the steps of:
clustering the historical data set to be processed according to a preset clustering model, and extracting target historical data from the clustered historical data set to be processed;
and forming a simulated asset pool according to the target historical data, and predicting the real-time early compensation rate of the simulated asset pool.
Optionally, the step of forming a simulated asset pool according to the target historical dataset and predicting a real estate early compensation rate of the simulated asset pool includes:
merging the target historical data to form a simulated asset pool, and determining an early compensation rate curve of the simulated asset pool;
predicting the house loan early compensation rate according to the early compensation rate curve;
wherein, the early compensation rate of the house loan is:
wherein P represents the early compensation amount in the month, C represents the uncompensated principal amount in the month, and M represents the principal to be returned in the month.
Optionally, the step of clustering the to-be-processed historical data set according to a preset clustering model and extracting target historical data from the clustered to-be-processed historical data set includes:
dividing the historical data set to be processed into different types of historical data according to a preset clustering model;
acquiring house loan data which are the same as the historical data of different categories from an asset pool to be predicted, and dividing the house loan data into corresponding categories;
and extracting target historical data from the classified different types of historical data.
Optionally, the step of extracting the target historical data from the classified different types of historical data includes:
labeling the divided historical data of different categories;
dividing the divided different types of historical data into first historical data and second historical data according to labels;
deleting the first historical data, and extracting target historical data from the second historical data according to the quantity of the first historical data.
Optionally, before the step of clustering the to-be-processed historical data set according to a preset clustering model and extracting the target historical data from the clustered to-be-processed historical data set, the method further includes:
and associating the historical data to be processed with the feature wide table, and constructing a preset clustering model according to the information of the feature wide table after association.
Optionally, before the step of clustering the to-be-processed historical data set according to a preset clustering model and extracting the target historical data from the clustered to-be-processed historical data set, the method further includes:
determining a simulation date of the real-time loan early compensation rate according to the prediction date of the real-time loan early compensation rate, and determining a slice historical data set based on the simulation date;
and carrying out data layering and data screening on the slice historical data set to obtain a historical data set to be processed.
Optionally, the step of performing data layering and data screening on the slice historical dataset to obtain a historical dataset to be processed includes:
data layering is carried out on the slice historical data set, and the layered slice historical data set is classified according to the dimension parameter, the regularization parameter and the sparsity parameter;
clustering is carried out according to the classification result, the membership matrix and the fuzzy mean value, and a slice historical dataset is screened according to the clustering result;
when the screening times do not reach the preset times, fixing the sparsity parameters and the clustering center and updating the screened slice historical data set;
or when the screening parameters reach the preset times, obtaining a historical data set to be processed.
In addition, in order to achieve the above object, the present invention also provides a house loan early-compensation rate predicting device, comprising: the system comprises an acquisition module and a prediction module;
the acquisition module is used for clustering the historical data set to be processed according to a preset clustering model and extracting target historical data from the clustered historical data set to be processed;
and the prediction module is used for forming a simulated asset pool according to the target historical data and predicting the real-time loan early-compensation rate of the simulated asset pool.
In addition, in order to achieve the above object, the present invention also proposes a loan early-compensation rate prediction apparatus, which includes a memory, a processor, and a loan early-compensation rate prediction program stored on the memory and executable on the processor, the loan early-compensation rate prediction program being configured to implement the loan early-compensation rate prediction method as described above.
In addition, in order to achieve the above object, the present invention also proposes a storage medium having stored thereon a loan early-compensation rate prediction program which, when executed by a processor, implements the loan early-compensation rate prediction method as described above.
The invention discloses a house loan early compensation rate prediction method, a device, equipment and a storage medium, wherein the method comprises the following steps: clustering the historical data set to be processed according to a preset clustering model, and extracting target historical data from the clustered historical data set to be processed; and forming a simulated asset pool according to the target historical data, and predicting the house loan early compensation rate of the simulated asset pool. According to the invention, the historical data sets to be processed are clustered according to the preset clustering model, the data of the same type are divided into the same type, the historical data sets to be processed are sampled according to the divided types to obtain the target historical data sets, a simulated asset pool is formed according to the target historical data sets, and the early compensation rate of the house and the loan is predicted based on the simulated asset pool, so that the early compensation rate is not required to be fixed in a certain interval according to expert experience, and can not be timely adjusted according to market trend, and the prediction accuracy of the early compensation rate of the house and the loan is improved.
Drawings
FIG. 1 is a schematic diagram of a construction of a loan early-compensation rate prediction device for a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a flow chart of a first embodiment of a method for predicting a real-time early compensation rate of a house according to the present invention;
FIG. 3 is a flow chart of data screening according to an embodiment of the present invention;
FIG. 4 is a general flow chart of the prediction of the early rate of the house loan according to one embodiment of the present invention;
FIG. 5 is a flowchart of a second embodiment of a method for predicting a real-time early compensation rate of a house;
FIG. 6 is a flowchart of a third embodiment of a method for predicting a premature rate of a house credit according to the invention;
FIG. 7 is a block diagram showing the construction of a first embodiment of a device for predicting a premature rate of a house.
The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a loan early-compensation rate prediction device of a hardware operation environment according to an embodiment of the present invention.
As shown in fig. 1, the loan early-compensation rate prediction apparatus may include: a processor 1001, such as a central processing unit (Central Processing Unit, CPU), a communication bus 1002, a user interface 1003, a network interface 1004, a memory 1005. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display (Display), and the optional user interface 1003 may also include a standard wired interface, a wireless interface, and the wired interface for the user interface 1003 may be a USB interface in the present invention. The network interface 1004 may optionally include a standard wired interface, a Wireless interface (e.g., a Wireless-Fidelity (Wi-Fi) interface). The Memory 1005 may be a high-speed random access Memory (Random Access Memory, RAM) or a stable Memory (NVM), such as a disk Memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.
It will be appreciated by those skilled in the art that the structure shown in fig. 1 does not constitute a limitation of the loan early compensation rate prediction apparatus, and may include more or fewer components than shown, or may combine certain components, or may be arranged in a different arrangement of components.
As shown in FIG. 1, memory 1005, which is considered a computer storage medium, may include an operating system, a network communication module, a user interface module, and a loan early compensation rate prediction program.
In the equipment for predicting the early compensation rate of the house, shown in fig. 1, the network interface 1004 is mainly used for connecting a background server and performing data communication with the background server; the user interface 1003 is mainly used for connecting user equipment; the loan early compensation rate prediction device invokes a loan early compensation rate prediction program stored in the memory 1005 through the processor 1001, and executes the loan early compensation rate prediction method provided by the embodiment of the invention.
Based on the hardware structure, the embodiment of the house loan early compensation rate prediction method is provided.
Referring to fig. 2, fig. 2 is a flowchart illustrating a first embodiment of a method for predicting a premature rate of a house, according to the present invention.
Step S10: clustering the historical data set to be processed according to a preset clustering model, and extracting target historical data from the clustered historical data set to be processed.
It should be noted that, the execution body of the present embodiment may be a computer software device having functions of data processing, network communication, and program running, for example, a loan early compensation rate prediction device, or other electronic devices capable of implementing the same or similar functions, which is not limited in this embodiment.
It should be appreciated that, at present, expert experience fixes the predicted early compensation rate at 8% or 10%, and there are significant limitations to fixing early compensation rate, such as: 1) The fixed early compensation rate adopts expert experience values, lacks multidimensional information support and has larger error; 2) The adoption of the fixed early compensation rate has lower flexibility, can not reflect the market trend of the house loan early compensation rate, and influences the security assessment accuracy.
In order to overcome the defects, aiming at the problem of measuring and calculating the early compensation rate of key factors affecting the pricing of the house mortgage securities, the embodiment adopts an algorithm based on KMeas clustering to construct a simulated asset pool approximate to the predicted asset pool, and uses the historical early compensation rate curve of the simulated asset pool to calculate the early compensation rate curve of the predicted asset pool.
It can be appreciated that in this embodiment, a KMeans clustering model is established by using the multidimensional feature of the housing loan, similar "to-be-predicted property pool" data and historical house loan data are divided into the same category, and then, historical data are extracted from different categories to construct a similar "simulated property pool" as the "to-be-predicted property pool"; the real early compensation rate curve of the lifetime of the asset pool can be calculated by using the historical early compensation of the simulated asset pool. Compared with the fixed early compensation value in the prior art, the fluctuation early compensation rate curve predicted by the embodiment fuses multidimensional information of the house loan, can more accurately reflect the expected quotation of the house loan market repayment in advance, and improves the security of the redemption of the security product and the accuracy of expected income.
It is understood that the historical data set to be processed may be identity information of the borrower, loan information of the borrower, repayment information of the borrower, and the like.
It should be noted that, clustering the historical data set to be processed may be performed by Kmeans clustering, where data of the same category is divided into the same category, and hierarchical clustering is performed on the historical data to be processed, that is, kmeans clustering is performed on samples divided into different payout sets respectively.
It can be understood that the target historical data is obtained by sampling the historical data to be processed in each category by a hierarchical sampling method.
Further, in order to improve the prediction accuracy of the early compensation rate of the house, the method further includes, before step S10:
and associating the historical data to be processed with the feature wide table, and constructing a preset clustering model according to the information of the feature wide table after association.
The feature width table may be constituted by borrower information, repayment details, loan information, and account information. The borrower information includes fields of birthday month ', ' sex ', ' place of home book ', ' marital status ', ' education level ', ' degree of education ', etc. The loan information includes fields of year of release ', ' month of release ', ' amount of release ', ' total number of months of release ', ' line of business ', ' institution ', ' agency ', and the like. The repayment details comprise fields such as 'principal should be returned', 'early compensation amount', 'accumulated early compensation amount', and the like. The account information includes fields of 'principal balance', 'lending ten-level classification code', 'account age', and the like.
It can be understood that the to-be-processed historical data is related to the feature width table, and a Kmeans cluster model (preset cluster model) is built by using the borrower information, repayment detail, loan information and account information in the feature width table.
Further, in order to improve the prediction accuracy of the early compensation rate of the house, the method further includes, before step S10:
determining a simulation date of the real-time loan early compensation rate according to the prediction date of the real-time loan early compensation rate, and determining a slice historical data set based on the simulation date;
and carrying out data layering and data screening on the slice historical data set to obtain a historical data set to be processed. It should be noted that "simulation date" is determined based on the predicted start date "Wherein->To round up, the predicted start date T1, the predicted period number T. The simulated date calculation method can ensure that the simulated date of the simulated asset pool is the same month as the measured initial date of the RMBS asset pool to be measured and calculated, and the influence of special months and seasons on the early compensation rate is avoided.
It is to be understood that the loan data whose update date is equal to the simulation date is selected from the full-volume historical data set as the slice historical data set.
It can be understood that the periodic fluctuation rules of the early compensation curves of different paying-off intervals are different, and the intervals of peaks and valleys of the early compensation rate have larger differences. If the loans with different payouts are processed together, the periodic characteristics of the early compensation curves of different payouts can be blurred. Therefore, the present embodiment firstly divides the data samples into different intervals according to different loan release amounts, forming a plurality of sets of 0-10 ten thousand, 10-20 ten thousand, 20-30 ten thousand, 30-40 ten thousand, 40-50 ten thousand, 50-60 ten thousand, 60-70 ten thousand, 70-80 ten thousand, 80-90 ten thousand, 90-100 ten thousand, 100-110 ten thousand and more than 110, and each interval is respectively subjected to subsequent operations.
It will be appreciated that since the samples in the issued RMBS asset pool are screened for house loan data, the screening conditions are standards established on the pool contract. Therefore, when the samples of the simulated asset pool are screened, the screening conditions and the self account age distribution characteristics of the asset pool to be predicted are referred to at the same time, and the screening rules are as follows: 1) The ten-level classification is from normal level to five-level; (but the attention level is found in the pool of assets already released, and therefore added when historical data is later selected); 2) The current period cannot be the clearing date or the 0 th period; 3) The loan product code must be a hand or second room; 4) The loan period is less than 30 years, and the loan amount is less than 1200 ten thousand (when the loan is issued); 5) The sum of the rest period of the loan and the age at the time is less than 70; 6) The repayment mode is to pay off-equal cost according to month or pay off-equal cost according to month; 7) The overdue principal balance is 0, and the overdue accumulated occurrence number < =2; 8) The current repayment state must be normal and repaid; 9) And (3) referring to the average account age of the asset pool data to be predicted in each set, and removing samples with too large or too small account ages.
It will be appreciated that after the slice history data set is screened, preprocessing operations such as cleaning, missing value processing, outlier processing, feature encoding, etc. are further required to be performed on the slice history data set.
Further, in order to improve the prediction accuracy of the early compensation rate of the house loan, the step of performing data layering and data screening on the slice historical data set to obtain a historical data set to be processed includes:
data layering is carried out on the slice historical data set, and the layered slice historical data set is classified according to the dimension parameter, the regularization parameter and the sparsity parameter;
clustering is carried out according to the classification result, the membership matrix and the fuzzy mean value, and a slice historical dataset is screened according to the clustering result;
when the screening times do not reach the preset times, fixing the sparsity parameters and the clustering center and updating the screened slice historical data set;
or when the screening parameters reach the preset times, obtaining a historical data set to be processed.
It should be noted that, in this embodiment, the membership degree pair with the sparse constraint is utilized to select ring, such as ambiguity, so as to adaptively learn the sparse membership degree, thereby obtaining the discrimination projection matrix (to-be-processed historical data set) in the low latitude space, and the feature of discrimination is selected by simultaneously performing the ambiguity learning and the sparse learning.
For easy understanding, referring to fig. 3, fig. 3 is a data filtering flow chart, in which a slice history data set is input, parameters c (dimensional parameters), α (regularization parameters) and γ (sparsity parameters) are used to classify the layered slice history data set, obtain a sample class 1, a sample class 2, a term, and a sample class 3, introduce a membership matrix Y and cluster according to a fuzzy k-means, calculate an objective function, calculate W (filtering slice history data set), and fix parameters Y, m when the filtering times do not reach the running times (preset times) j And (a clustering center) returning to input the slice historical data set, classifying the layered slice historical data set by parameters c (dimension parameters), alpha (regularization parameters) and gamma (sparsity parameters), updating the step W until the screening times reach the operation times, and outputting the W.
Step S20: and forming a simulated asset pool according to the target historical data, and predicting the real-time early compensation rate of the simulated asset pool.
It should be noted that, in this embodiment, the input "to-be-predicted asset pool" account number and the two-hand house credit historical data sample are firstly divided into different sections according to the different payouts, and then the subsequent modeling is performed to preserve the periodicity of early compensation rates in the different payouts. And then, screening house loan history samples in different loan amount intervals according to the pool entering standard and account age distribution of the asset pool to be predicted. And then, carrying out KMeans clustering modeling on the data in different paying intervals according to the multidimensional loan information, sampling the historical data of each category in the different paying intervals, and merging to form a simulated property pool similar to the property pool to be predicted. And finally, calculating a historical early compensation rate curve of the simulated asset pool, and taking the curve as an early compensation rate curve in a time period required to be calculated by the asset pool to be predicted.
For easy understanding, referring to fig. 4, fig. 4 is a general flow chart of early compensation rate prediction for a house, first performing data acquisition and processing, determining a simulation date according to a calculation start date, acquiring a data set with the same calculation start date and simulation date from a full-scale historical data set of a two-hand house as a slice historical data set, performing data layering based on an account number of a property pool to be predicted and a feature wide table (borrower information, repayment detail, loan information and account information), then performing simulated property pool construction, performing layered data screening according to layered data, performing data preprocessing based on the feature wide table and KMeans clustering on the data, performing layered sampling after clustering is completed, combining the data after layered sampling to form a simulated property pool, finally performing early compensation rate curve prediction, and performing early compensation rate calculation based on the simulated property pool according to a balance wide table, an early compensation wide table and a calculation period number, thereby obtaining a predicted early compensation rate.
According to the embodiment, a historical data set to be processed is clustered according to a preset clustering model, and target historical data is extracted from the clustered historical data set to be processed; and forming a simulated asset pool according to the target historical data, and predicting the house loan early compensation rate of the simulated asset pool. According to the embodiment, the historical data sets to be processed are clustered according to the preset clustering model, the same type of data is divided into the same type, the historical data sets to be processed are sampled according to the divided types to obtain the target historical data sets, a simulated asset pool is formed according to the target historical data sets, and the early compensation rate of the house loan is predicted based on the simulated asset pool, so that the early compensation rate is not required to be fixed in a certain interval according to expert experience, and cannot be timely adjusted according to market trend, and the prediction accuracy of the early compensation rate of the house loan is improved.
Referring to fig. 5, fig. 5 is a flowchart illustrating a second embodiment of the method for predicting a real-time early-compensation rate of a house according to the present invention, and the second embodiment of the method for predicting a real-time early-compensation rate of a house according to the present invention is proposed based on the first embodiment shown in fig. 2.
In a second embodiment, the step S20 includes:
step S201: merging the target historical data to form a simulated asset pool, and determining an early compensation rate curve of the simulated asset pool;
step S202: and predicting the house loan early compensation rate according to the early compensation rate curve.
It can be understood that the target historical data are combined to form a simulated asset pool, and finally, a historical early compensation rate curve of the simulated asset pool is calculated and is used as an early compensation rate curve in a time period which needs to be measured and calculated by the asset pool to be predicted.
When calculating the historical early compensation rate, the general calculation formula of the early compensation rate of the asset pool of the securitization project of the credit asset according to the market practice is as follows:
wherein P represents the early compensation amount in the month, C represents the uncompensated principal amount in the month, and M represents the principal to be returned in the month.
It will be understood that this refers to modeling principal balances of all accounts in the asset pool at the end of the last month, i.e., the beginning of the month, and that M is the principal that is planned to be refunded in the month, and this refers to modeling principal that all accounts in the asset pool are expected to be refunded in the month. In practical operation, considering the convenience of data acquisition, the balance C of the outstanding principal in the beginning of the month is generally directly adopted in denominator calculation, and the principal is not deducted from the plan of the month. Thus, the above formula can also be expressed as:
it should be noted that this adjustment has little practical impact on the early compensation rate data and can be ignored. According to the formula, the early compensation wide table and the balance wide table of all the historical loan early compensation data and the early balance data of the month can be used for calculating the historical month early compensation rate of the simulated property pool for T months after the simulated date T2, wherein the early compensation rate is the prediction result of the property pool to be predicted.
The embodiment combines the target historical data to form a simulated asset pool, and determines an early compensation rate curve of the simulated asset pool; and predicting the house loan early compensation rate according to the early compensation rate curve. The embodiment constructs a simulated asset pool similar to the asset pool to be predicted, predicts the early compensation rate based on the simulated asset pool, and takes the prediction result as the prediction result of the asset pool to be predicted, thereby improving the flexibility and the accuracy of the early compensation rate of the house loan.
Referring to fig. 6, fig. 6 is a flowchart illustrating a third embodiment of the method for predicting a real-time early-compensation rate of a house according to the present invention, and the third embodiment of the method for predicting a real-time early-compensation rate of a house according to the present invention is proposed based on the first embodiment shown in fig. 2.
In a third embodiment, the step S10 further includes:
step S101: dividing the historical data set to be processed into different types of historical data according to a preset clustering model;
step S102: acquiring house loan data which are the same as the historical data of different categories from an asset pool to be predicted, and dividing the house loan data into corresponding categories;
step S103: and extracting target historical data from the classified different types of historical data.
It will be appreciated that the data in the pool of assets to be predicted may be divided into the same data categories as their corresponding historical data sets to be processed.
Further, in order to improve the prediction accuracy of the early compensation rate of the house, step S103 of this embodiment may include:
labeling the divided historical data of different categories;
dividing the divided different types of historical data into first historical data and second historical data according to labels;
deleting the first historical data, and extracting target historical data from the second historical data according to the quantity of the first historical data.
It should be noted that, the first historical data is asset pool data, and the second historical data is non-asset pool historical data.
In the specific implementation, labeling each sample according to the category generated by clustering in different payoff amount intervals, removing asset pool data in each category, leaving non-asset pool historical data, and randomly sampling the same amount of simulation data in the non-asset pool historical data according to the amount of the asset pool data.
According to the embodiment, the historical data set to be processed is divided into different types of historical data according to a preset clustering model; acquiring house loan data which are the same as the historical data of different categories from an asset pool to be predicted, and dividing the house loan data into corresponding categories; and extracting target historical data from the classified different types of historical data. According to the embodiment, the historical data set to be processed is divided into the historical data of different categories according to the preset clustering model, the house loan data of the same category is obtained from the asset pool to be predicted and divided into the same category, and the historical data is extracted from each category, so that the data of the simulated asset pool is more accurate, and the prediction accuracy of the house loan early compensation rate is improved.
In addition, the embodiment of the invention also provides a storage medium, wherein the storage medium is stored with a house loan early compensation rate prediction program, and the house loan early compensation rate prediction program realizes the house loan early compensation rate prediction method when being executed by a processor.
In addition, referring to fig. 7, an embodiment of the present invention further provides a device for predicting a premature rate of a house, where the device for predicting a premature rate of a house includes: an acquisition module 10 and a prediction module 20;
the acquiring module 10 is configured to cluster the historical data set to be processed according to a preset cluster model, and extract target historical data from the clustered historical data set to be processed;
the prediction module 20 is configured to form a simulated asset pool according to the target historical data, and predict a real estate early compensation rate of the simulated asset pool.
According to the embodiment, a historical data set to be processed is clustered according to a preset clustering model, and target historical data is extracted from the clustered historical data set to be processed; and forming a simulated asset pool according to the target historical data, and predicting the house loan early compensation rate of the simulated asset pool. According to the embodiment, the historical data sets to be processed are clustered according to the preset clustering model, the same type of data is divided into the same type, the historical data sets to be processed are sampled according to the divided types to obtain the target historical data sets, a simulated asset pool is formed according to the target historical data sets, and the early compensation rate of the house loan is predicted based on the simulated asset pool, so that the early compensation rate is not required to be fixed in a certain interval according to expert experience, and cannot be timely adjusted according to market trend, and the prediction accuracy of the early compensation rate of the house loan is improved.
Based on the first embodiment of the device for predicting the early compensation rate of the house loan according to the present invention, a second embodiment of the device for predicting the early compensation rate of the house loan according to the present invention is provided.
In this embodiment, the prediction module 20 is configured to combine the target history data to form a simulated asset pool, and determine an early-compensation rate curve of the simulated asset pool.
Further, the prediction module 20 is further configured to predict a house loan early-compensation rate according to the early-compensation rate curve.
Further, the obtaining module 10 is further configured to divide the to-be-processed historical data set into different types of historical data according to a preset clustering model.
Further, the obtaining module 10 is further configured to obtain, from the pool of assets to be predicted, the same house loan data as the historical data of the different categories, and divide the house loan data into the corresponding categories.
Further, the obtaining module 10 is further configured to extract the target historical data from the classified different types of historical data.
Further, the obtaining module 10 is further configured to label the classified historical data of different categories.
Further, the obtaining module 10 is further configured to divide the divided different types of history data into a first history data and a second history data according to a label.
Further, the obtaining module 10 is further configured to delete the first history data, and extract target history data from the second history data according to the number of the first history data.
Further, the obtaining module 10 is further configured to correlate the data to be processed with the feature broad table, and construct a preset clustering model according to the information of the correlated feature broad table.
Further, the obtaining module 10 is further configured to determine a simulation date of the real-time compensation rate according to the prediction date of the real-time compensation rate, and determine a slice history data set based on the simulation date.
Further, the obtaining module 10 is further configured to perform data layering and data screening on the slice historical dataset to obtain a historical dataset to be processed.
Further, the obtaining module 10 is further configured to perform data layering on the slice history data set, and classify the layered slice history data set according to the dimension parameter, the regularization parameter and the sparsity parameter.
Further, the obtaining module 10 is further configured to cluster according to the classification result, the membership matrix and the fuzzy mean value, and screen the slice history dataset according to the clustering result.
Further, the obtaining module 10 is further configured to fix the sparsity parameter and the clustering center and update the post-screening slice history dataset when the screening number does not reach the preset number.
Other embodiments or specific implementations of the device for predicting a real-time early-compensation rate of a house may refer to the above method embodiments, and will not be described herein.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. read only memory mirror (Read Only Memory image, ROM)/random access memory (Random Access Memory, RAM), magnetic disk, optical disk), comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the method according to the embodiments of the present invention.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.
Claims (10)
1. The house loan early compensation rate prediction method is characterized by comprising the following steps of:
clustering the historical data set to be processed according to a preset clustering model, and extracting target historical data from the clustered historical data set to be processed;
and forming a simulated asset pool according to the target historical data, and predicting the real-time early compensation rate of the simulated asset pool.
2. The method of claim 1, wherein the step of forming a simulated pool of assets from the target historical dataset and predicting the real estate early rate of the simulated pool of assets comprises:
merging the target historical data to form a simulated asset pool, and determining an early compensation rate curve of the simulated asset pool;
predicting the house loan early compensation rate according to the early compensation rate curve;
wherein, the early compensation rate of the house loan is:
wherein P represents the early compensation amount in the month, C represents the uncompensated principal amount in the month, and M represents the principal to be returned in the month.
3. The house loan early-compensation rate predicting method of claim 1, wherein the step of clustering the to-be-processed history data sets according to a preset clustering model and extracting target history data from the clustered to-be-processed history data sets comprises:
dividing the historical data set to be processed into different types of historical data according to a preset clustering model;
acquiring house loan data which are the same as the historical data of different categories from an asset pool to be predicted, and dividing the house loan data into corresponding categories;
and extracting target historical data from the classified different types of historical data.
4. The method of claim 3, wherein the step of extracting the target history data from the divided different categories of history data comprises:
labeling the divided historical data of different categories;
dividing the divided different types of historical data into first historical data and second historical data according to labels;
deleting the first historical data, and extracting target historical data from the second historical data according to the quantity of the first historical data.
5. The method for predicting a real estate asset early compensation rate of claim 1 wherein the step of clustering the historical data sets to be processed according to a preset clustering model and extracting the target historical data from the clustered historical data sets to be processed further comprises:
and associating the historical data to be processed with the feature wide table, and constructing a preset clustering model according to the information of the feature wide table after association.
6. The method for predicting a real estate asset early compensation rate of claim 1 wherein the step of clustering the historical data sets to be processed according to a preset clustering model and extracting the target historical data from the clustered historical data sets to be processed further comprises:
determining a simulation date of the real-time loan early compensation rate according to the prediction date of the real-time loan early compensation rate, and determining a slice historical data set based on the simulation date;
and carrying out data layering and data screening on the slice historical data set to obtain a historical data set to be processed.
7. The method of claim 6, wherein the step of data layering and data screening the slice history data set to obtain a history data set to be processed comprises:
data layering is carried out on the slice historical data set, and the layered slice historical data set is classified according to the dimension parameter, the regularization parameter and the sparsity parameter;
clustering is carried out according to the classification result, the membership matrix and the fuzzy mean value, and a slice historical dataset is screened according to the clustering result;
when the screening times do not reach the preset times, fixing the sparsity parameters and the clustering center and updating the screened slice historical data set;
or when the screening parameters reach the preset times, obtaining a historical data set to be processed.
8. A house loan early compensation rate predicting device, characterized in that the house loan early compensation rate predicting device comprises: the system comprises an acquisition module and a prediction module;
the acquisition module is used for clustering the historical data set to be processed according to a preset clustering model and extracting target historical data from the clustered historical data set to be processed;
and the prediction module is used for forming a simulated asset pool according to the target historical data and predicting the real-time loan early-compensation rate of the simulated asset pool.
9. A house loan early-compensation rate predicting apparatus, characterized in that the house loan early-compensation rate predicting apparatus comprises: a memory, a processor and a loan early-compensation rate prediction program stored on the memory and executable on the processor, which, when executed by the processor, implements the steps of the loan early-compensation rate prediction method of any one of claims 1 to 7.
10. A storage medium having stored thereon a loan early-compensation rate prediction program which, when executed by a processor, implements the steps of the loan early-compensation rate prediction method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311208866.7A CN117196827A (en) | 2023-09-18 | 2023-09-18 | Method, device, equipment and storage medium for predicting early compensation rate of house loan |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311208866.7A CN117196827A (en) | 2023-09-18 | 2023-09-18 | Method, device, equipment and storage medium for predicting early compensation rate of house loan |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117196827A true CN117196827A (en) | 2023-12-08 |
Family
ID=89005016
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311208866.7A Pending CN117196827A (en) | 2023-09-18 | 2023-09-18 | Method, device, equipment and storage medium for predicting early compensation rate of house loan |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117196827A (en) |
-
2023
- 2023-09-18 CN CN202311208866.7A patent/CN117196827A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110415119B (en) | Model training method, bill transaction prediction method, model training device, bill transaction prediction device, storage medium and equipment | |
| CN109389494B (en) | Loan fraud detection model training method, loan fraud detection method and device | |
| US20150088783A1 (en) | System and method for modeling and quantifying regulatory capital, key risk indicators, probability of default, exposure at default, loss given default, liquidity ratios, and value at risk, within the areas of asset liability management, credit risk, market risk, operational risk, and liquidity risk for banks | |
| CN112116245A (en) | Credit risk assessment method, credit risk assessment device, computer equipment and storage medium | |
| CN110659961A (en) | Method and device for identifying off-line commercial tenant | |
| CN110363642A (en) | Loan data processing method, device, readable storage medium storing program for executing and program product | |
| CN112990386A (en) | User value clustering method and device, computer equipment and storage medium | |
| CN116596659A (en) | Enterprise intelligent credit approval method, system and medium based on big data wind control | |
| CN116777591A (en) | Training method of repayment capability prediction model, repayment capability prediction method and repayment capability prediction device | |
| Breeden et al. | Multihorizon discrete time survival models | |
| CN111626855A (en) | Bond credit interest difference prediction method and system | |
| US20210049687A1 (en) | Systems and methods of generating resource allocation insights based on datasets | |
| CN112561691B (en) | Client trust prediction method, device, equipment and storage medium | |
| CN116800831B (en) | Service data pushing method, device, storage medium and processor | |
| CN113807943A (en) | Multi-factor valuation method, system, medium and equipment for bad assets | |
| CN117132317A (en) | Data processing method, device, equipment, medium and product | |
| CN115829722A (en) | Training method of credit risk scoring model and credit risk scoring method | |
| CN117252677A (en) | Credit line determination method and device, electronic equipment and storage medium | |
| US20150170269A1 (en) | Real estate market condition indicator | |
| CN117196827A (en) | Method, device, equipment and storage medium for predicting early compensation rate of house loan | |
| Bulut et al. | Financial Failure Estimation with Logistic Regression Model: A Study on Technology Sector Companies Treated in BIST | |
| CN114418776A (en) | Data processing method, device, terminal equipment and medium | |
| CN113706258A (en) | Product recommendation method, device, equipment and storage medium based on combined model | |
| CN113177733A (en) | Medium and small micro-enterprise data modeling method and system based on convolutional neural network | |
| CN112785204A (en) | Method for determining object risk value and method for adjusting business parameter |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |