CN104778622A - Method and system for predicting TPS transaction event threshold value - Google Patents
Method and system for predicting TPS transaction event threshold value Download PDFInfo
- Publication number
- CN104778622A CN104778622A CN201510213263.5A CN201510213263A CN104778622A CN 104778622 A CN104778622 A CN 104778622A CN 201510213263 A CN201510213263 A CN 201510213263A CN 104778622 A CN104778622 A CN 104778622A
- Authority
- CN
- China
- Prior art keywords
- tps
- transaction
- data
- threshold value
- transaction event
- 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
Abstract
The invention discloses a method and system for predicting a TPS transaction event threshold value. The method comprises the following steps: acquiring transaction data, and extracting TPS data and TPS data features; conducting time series division according to time windows so that statistics can be conducted on the maximum historical transaction value and the minimum historical transaction value corresponding to each time window; training a data mining model so that a TPS transaction event prediction regression model can be obtained; according to the TPS transaction event prediction regression model and the maximum historical transaction value and the minimum historical transaction value corresponding to each time window, predicting the TPS transaction event threshold value. According to the prediction method, the TPS transaction event threshold value can be predicted, reference can be provided for improving backstage service, suggestions are also given for making decisions on trouble removal, and the use experience of a user is improved.
Description
Technical field
The present invention relates to Computer Applied Technology field, particularly the Forecasting Methodology of a kind of bank backstage TPS transaction event threshold value and prognoses system.
Background technology
Data mining is a kind of new business information treatment technology, is first applied to the field such as finance, telecommunications, and principal feature extracts mass data, changes, analyzes and modelling process, therefrom extracts the critical data contributing to business decision.
Developing rapidly of such as Bank Informatization, thus create a large amount of business datums, so extract valuable information from mass data, thus being the business decision service of bank, is the important applied field of data mining.Present stage, the application of data mining in banking industry, mainly can be divided into the following aspects.1) risk.Data mining is risk management in one of the important application of banking industry, as assessing credit risks.By building credit Rating Model, the risk of assessment creditor or credit card application people.A solution of carrying out assessing credit risks, can formulate credit rating standard to accounts all in banking data base, just can draw the list of credit risk with some data base queryings.This for the low-risk grading of height or classification, be the account features based on each client, as still outstanding loan, credit downgrade report record, Account Type, income level and other information etc.By data mining, abnormal credit card service condition can also be obtained, determine the consumer behavior of extreme client.According to historical data, evaluation causes the characteristic sum background of credit risk client, the client of risk of loss may be caused, to in the standing of client and the basis of management forecast, the method of utilization system identifies the type of credit risk and reason, evaluates and tests, find the inducement causing credit risk, effectively control and reduce the generation of credit risk.2) fault pre-alarming.As large-scale financial institution, its security and high efficiency just seem particularly important, wherein the lifeblood of security banking system especially, but even so, the large-scale fault in bank aspect still happens occasionally.And large-scale fault is not often caused by the work mistake on foreground, because the thorough transaction step in bank foreground almost can stop the generation of human error, even and if error occurs also to be the small-scale mistake of one or two transaction.Large-scale fault is all often caused by the fault of the system on backstage.Therefore, want the generation more effectively avoiding bank's fault, should set about from background system emphatically.But bank's background system is often very complicated, cause the reason of fault varied especially, may be due to: the linked network between bank, the database of rear end record data, produces fault for server running transaction program etc.And one of them fault often causes a series of chain reaction, such as, when paralysis occurs database, all transaction request will start to pile up, thus cause the inadequate resource of server; On the contrary, leak if the internal memory of server produces, system resource so gradually can be fewer and feweri, thus cause the operation resource requirement of database not enough, finally paralyses.
As can be seen here, the system coherence on bank backstage is quite complicated, want by rule and method Direct Analysis be out of order Producing reason hardly may.But, although the number of times that fault produces is rare, but be not irregular following, according to the experience of bank aspect, often system can produce the state of some exceptions before the failure occurs, and the state of system is often more prone to monitoring than fault, so can by the parameter of real-time monitoring analysis system, thus when prediction fault will occur, wherein TPS (Transaction Per Second, issued transaction amount per second) not only contributes to bank's failure prediction but also recovers to play an important role for the bank backstage after breaking down.Therefore, how to predict that TPS transaction event threshold value seems particularly important.
Summary of the invention
The present invention is intended to solve one of technical matters in above-mentioned correlation technique at least to a certain extent.
For this reason, one object of the present invention is to propose a kind of Forecasting Methodology can predicting the TPS transaction event threshold value of TPS transaction event threshold value.
Another object of the present invention is the prognoses system proposing a kind of TPS transaction event threshold value.
For achieving the above object, one aspect of the present invention embodiment proposes a kind of Forecasting Methodology of TPS transaction event threshold value, comprises the following steps: obtain transaction data, and from described transaction data, extract issued transaction amount TPS data per second and TPS data characteristics; According to time window, time series segmentation is carried out to described TPS data, to add up historical trading maximal value corresponding to each time window and minimum value; According to described TPS data characteristics training data mining model to obtain TPS transaction event prediction regression model; And according to described TPS transaction event prediction regression model and historical trading maximal value corresponding to each time window described and minimum value prediction TPS transaction event threshold value.
According to the Forecasting Methodology of the TPS transaction event threshold value that the embodiment of the present invention proposes, by extracting TPS data and TPS data characteristics from transaction data, to add up historical trading maximal value corresponding to each time window and minimum value and to obtain TPS transaction event prediction regression model, thus prediction TPS transaction event threshold value, improve the reliability that data are extracted, realize the object of prediction TPS transaction event threshold value, not only provide reference for background service improves, and suggestion is provided to the decision-making of failture evacuation, improve the experience of user, simple and convenient.
In addition, the Forecasting Methodology of TPS transaction event threshold value according to the above embodiment of the present invention can also have following additional technical characteristic:
Further, in one embodiment of the invention, said method also comprises: show TPS transaction event threshold value by image conversion mode and predict the outcome.
Further, in one embodiment of the invention, select described data mining model according to type of transaction, wherein, described data mining model comprises k Neighborhood Model, BP neural network model or Random Forest model.
Further, in one embodiment of the invention, described TPS data characteristics is the feature that the multiple relevant information extracted from described transaction data connects each moment formed, wherein, described multiple relevant information comprises one or more in currency transaction information, current date week, in the past contemporaneous information and day trade transaction amplification information.
Further, in one embodiment of the invention, from described transaction data, TPS data characteristics is extracted by principal component analysis (PCA) or Method for Feature Selection.
The present invention on the other hand embodiment proposes a kind of prognoses system of TPS transaction event threshold value, comprising: data preprocessing module, for obtaining transaction data, and from described transaction data, extracting TPS data and TPS data characteristics; Statistical module, for carrying out time series segmentation according to time window to described TPS data, to add up historical trading maximal value corresponding to each time window and minimum value; Regression block, for predicting regression model according to described TPS data characteristics training data mining model to obtain TPS transaction event; And prediction module, for according to described TPS transaction event prediction regression model and historical trading maximal value corresponding to each time window described and minimum value prediction TPS transaction event threshold value.
According to the prognoses system of the TPS transaction event threshold value that the embodiment of the present invention proposes, by extracting TPS data and TPS data characteristics from transaction data, to add up historical trading maximal value corresponding to each time window and minimum value and to obtain TPS transaction event prediction regression model, thus prediction TPS transaction event threshold value, improve the reliability that data are extracted, realize the object of prediction TPS transaction event threshold value, not only provide reference for background service improves, and suggestion is provided to the decision-making of failture evacuation, improve the experience of user, simple and convenient.
In addition, the prognoses system of TPS transaction event threshold value according to the above embodiment of the present invention can also have following additional technical characteristic:
Further, in one embodiment of the invention, said system also comprises: graphic user interface, predicts the outcome for being shown TPS transaction event threshold value by image conversion mode.
Further, in one embodiment of the invention, described regression block is also for selecting described data mining model according to type of transaction, and wherein, described data mining model comprises k Neighborhood Model, BP neural network model or Random Forest model.
Further, in one embodiment of the invention, described TPS data characteristics is the feature that the multiple relevant information extracted from described transaction data connects each moment formed, wherein, described multiple relevant information comprises one or more in currency transaction information, current date week, in the past contemporaneous information and day trade transaction amplification information.
Further, in one embodiment of the invention, from described transaction data, TPS data characteristics is extracted by principal component analysis (PCA) or Method for Feature Selection.
The aspect that the present invention adds and advantage will part provide in the following description, and part will become obvious from the following description, or be recognized by practice of the present invention.
Accompanying drawing explanation
Above-mentioned and/or additional aspect of the present invention and advantage will become obvious and easy understand from accompanying drawing below combining to the description of embodiment, wherein:
Fig. 1 is the process flow diagram of the Forecasting Methodology of TPS transaction event threshold value according to the embodiment of the present invention;
Fig. 2 is the process flow diagram of the Forecasting Methodology of TPS transaction event threshold value according to an embodiment of the invention;
Fig. 3 is the process flow diagram of the Forecasting Methodology of TPS transaction event threshold value according to the present invention's specific embodiment;
Fig. 4 is the process flow diagram of TPS data characteristics extracting method according to an embodiment of the invention;
Fig. 5 to predict the outcome schematic diagram for TPS transaction event threshold value according to an embodiment of the invention;
Fig. 6 is the structural representation of the prognoses system of TPS transaction event threshold value according to the embodiment of the present invention; And
Fig. 7 is the structural representation given according to the prognoses system of the TPS transaction event threshold value of the present invention's specific embodiment.
Embodiment
Be described below in detail embodiments of the invention, the example of described embodiment is shown in the drawings, and wherein same or similar label represents same or similar element or has element that is identical or similar functions from start to finish.Be exemplary below by the embodiment be described with reference to the drawings, be intended to for explaining the present invention, and can not limitation of the present invention be interpreted as.
In addition, term " first ", " second " only for describing object, and can not be interpreted as instruction or hint relative importance or imply the quantity indicating indicated technical characteristic.Thus, be limited with " first ", the feature of " second " can express or impliedly comprise one or more these features.In describing the invention, the implication of " multiple " is two or more, unless otherwise expressly limited specifically.
In the present invention, unless otherwise clearly defined and limited, the term such as term " installation ", " being connected ", " connection ", " fixing " should be interpreted broadly, and such as, can be fixedly connected with, also can be removably connect, or connect integratedly; Can be mechanical connection, also can be electrical connection; Can be directly be connected, also indirectly can be connected by intermediary, can be the connection of two element internals.For the ordinary skill in the art, above-mentioned term concrete meaning in the present invention can be understood as the case may be.
In the present invention, unless otherwise clearly defined and limited, fisrt feature second feature it " on " or D score can comprise the first and second features and directly contact, also can comprise the first and second features and not be directly contact but by the other characterisation contact between them.And, fisrt feature second feature " on ", " top " and " above " comprise fisrt feature directly over second feature and oblique upper, or only represent that fisrt feature level height is higher than second feature.Fisrt feature second feature " under ", " below " and " below " comprise fisrt feature directly over second feature and oblique upper, or only represent that fisrt feature level height is less than second feature.
Describe Forecasting Methodology and the prognoses system of the TPS transaction event threshold value proposed according to the embodiment of the present invention with reference to the accompanying drawings, describe the Forecasting Methodology of the TPS transaction event threshold value proposed according to the embodiment of the present invention first with reference to the accompanying drawings.With reference to shown in Fig. 1, this Forecasting Methodology comprises the following steps:
S101, obtains transaction data, and from transaction data, extracts issued transaction amount TPS data per second and TPS data characteristics.
Wherein, in one embodiment of the invention, TPS data characteristics is the feature that the multiple relevant information extracted from transaction data connects each moment formed, wherein, multiple relevant information comprises one or more in currency transaction information, current date week, in the past contemporaneous information and day trade transaction amplification information.
Further, in one embodiment of the invention, from transaction data, TPS data characteristics is extracted by principal component analysis (PCA) or Method for Feature Selection.
S102, carries out time series segmentation according to time window to TPS data, to add up historical trading maximal value corresponding to each time window and minimum value.
S103, according to TPS data characteristics training data mining model to obtain TPS transaction event prediction regression model.
Wherein, in one embodiment of the invention, select data mining model according to type of transaction, wherein, data mining model comprises k Neighborhood Model, BP neural network model or Random Forest model.
S104, the historical trading maximal value corresponding with each time window according to TPS transaction event prediction regression model and minimum value prediction TPS transaction event threshold value.
Further, in one embodiment of the invention, the Forecasting Methodology of the embodiment of the present invention also comprises: show TPS transaction event threshold value by image conversion mode and predict the outcome.
In order to the Forecasting Methodology of the TPS transaction event threshold value proposed the embodiment of the present invention repeats better, be predicted as example with the TPS transaction event threshold value on bank backstage to be below described, it should be noted that the Forecasting Methodology of the embodiment of the present invention is not limited in a kind of application platform of bank.
In one particular embodiment of the present invention, as shown in Figure 2, the Forecasting Methodology of the embodiment of the present invention comprises the following steps:
S1, obtains bank's backstage transaction data, and extracts TPS data and feature thereof from bank's backstage transaction data, and wherein, TPS data characteristics refers to that the multiple relevant information extracted from bank's backstage transaction data links and the feature in each moment of formation.Wherein, multiple relevant information includes but not limited to: current trading situation, current date week, in the past the same period situation and day trade transaction amplification situation.
Particularly, as shown in Figure 3, the embodiment of the present invention can utilize feature extracting method to read in module by service data to obtain as the TPS data that were interval with 5 minutes from large scale business bank backstage transaction data, to carry out regression model training.TPS data are large scale business bank backstage All Activity and time of origin thereof mainly, by extracting useful information, such as: current trading situation, current date week, in the past the same period situation and day trade transaction amplification situation represent spaced features.The data set of TPS transaction event threshold value forecast model can be formed by this step.
In an embodiment of the present invention, as shown in Figure 4, the embodiment of the present invention can extract the TPS data characteristics in each moment from transaction data by feature extracting method.
Wherein, feature extracting method can comprise following several:
PCA (Principal Component Analysis, principal component analysis (PCA)), main thought is again projected on new coordinate system (major component) at data point, by maximize projection after data point between variance, optimization aim as shown by the following formula:
Wherein,
represent the data set after the impact of k dimension major component before deleting, W
krepresent a kth major component direction.
Thus, the maximum dimension of Data Placement dispersion can be found.Constantly repeat this process, find a major component at every turn, just the impact of this major component on data is temporarily eliminated, as shown by the following formula:
Wherein, X represents raw data set.
Like this, for the data space of D dimension, finally can finding out D projected dimensions, can know according to major component method for solving, dimensionality reduction work can be completed by suitably retaining several major components front.
Method for Feature Selection (Features Selection), does not change raw data set substantially, just therefrom extracts useful dimension subspace to complete dimensionality reduction.
And progressively forward saliency back-and-forth method is the simplest efficient a kind of feature selection approach, its main flow can illustrate by following step:
1, initial characteristics space is empty.
2, select a feature, the sorter of training after making to add feature space in the current situation can obtain the highest accuracy rate, and this feature is added feature space at every turn.
3, the 2nd step is repeated, until have selected the feature of enough dimensions.
4, trading volume maximal value and minimum value are added up.
S2, selects k Neighborhood Model, BP neural network model or Random Forest model, and according to TPS data training pattern to obtain the TPS transaction event threshold value prediction regression model of having trained.
Wherein, k Neighborhood Model is a kind of instance-based learning, or Local approximation and the study of the inertia after classification is postponed till in all calculating.During for regression model, object's property value can be assigned as the mean value of the property value of its k neighbour.BP neural network model is the mathematical model of a kind of mimic biology neural network structure and function.Neural network is calculated by a large amount of artificial neurons connection.By a large amount of nodes and between interconnect and form, a kind of specific output function of each node on behalf, be called excitation function, every two internodal connections all represent one for the weighted value by this connection signal, be referred to as weight, this is equivalent to the memory of neural network.The output of network then according to the bind mode of network, the difference of weighted value and excitation function and different.Random Forest model comprises multiple decision-tree model, and decision-tree model is obtained by following formula, and formula is:
P
lVar(Y
l)+P
rVar(Y
r)，
Wherein, P
lfor the leaf number of left subtree, P
rfor the leaf number of right subtree, Var represents and asks variance, Y
lfor all labels of left subtree, Y
rfor all labels of right subtree.
Specifically, if select k Neighborhood Model to carry out model training to training set.Training process is as follows:
1, first we determine k value (just refer to the k size of k near neighbor method, represent for a data point to be sorted, we will find its neighbour several) in advance
2, according to pre-determined distance metric formula, draw in the sample point of data point to be sorted and all known class, k nearest sample.
3, add up in this k sample, the quantity of each classification.
Random Forest model carries out model training for the training set of input.Random forest is a kind of integrated study model, is the comprehensive model of one combining the Bootstrap method of sampling, feature selecting, Bagging training method, decision-tree model.The C4.5 decision-tree model that decision-tree model is selected, and C4.5 decision tree is mainly used in classification problem, by the computing method of amendment classified information gain, just obtains a decision tree for regression forecasting thus.
The training of Random Forest model comprises following step:
1, for N number of original sample of input, employing is randomly drawed the mode put back to and is sampled, and obtains new N number of sample.
2, use the N number of sample training decision tree arrived of sampling, suppose the attribute that sample has M to tie up, so when node needs division, therefrom randomly draw out the attribute of M dimension, the rule according to C4.5 divides.
3, in the process of structure decision tree, each node needs to divide according to the rule of the 2nd step, a final formation decision tree.
4, constantly repeat 1 ~ 3 step, until obtain the decision tree needing number, just constitute random forest.
In the examples described above, because random forest is in fact the combination of many decision trees, therefore first decision-tree model is introduced.
Decision tree is that one forms disaggregated model by root step by step to leaf, the feature of each sensor selection problem to the best dimension of current division degree is classified to sample, concrete system of selection is different according to the version difference of decision tree, what adopt herein is the splitting method of C4.5 decision tree, specific as follows
Wherein, pi is the ratio belonging to i class in S, and A is the attribute of sample, and Values (A) is the codomain of attribute A, and Sv is the number of samples that in S, A attribute equals v.
Classical ID3 decision Tree algorithms uses information gain Gain (S, A) to assess and selects to divide attribute, but can find to use this evaluation index in using, and algorithm can be partial to the attribute selecting value more.In order to revise this shortcoming, Quilan proposes the improvement C4.5 algorithm of an ID3, uses information gain-ratio GainRatio (S, A) to choose division attribute, improves the accuracy of decision tree.
S3, to TPS transaction event threshold value prediction regression model input test collection data to carry out the prediction of TPS transaction event threshold value.
In an embodiment of the present invention, can find out that random forest is the aggregate of decision tree, when testing, as long as test respectively every decision tree, the result of test is voted by every decision tree and is obtained.
For evaluating the method for regression model, can according to the feature of commercial banks data: during rush periods, trading volume per second may have thousands of pen, and may only have two or three transaction in 5 minutes in the time-division in morning.Employ the mode that absolute error is combined with relative error to evaluate.Specifically when trading volume is lower than certain threshold value, we adopt absolute error to pass judgment on:
Δ＝X-L，
Wherein X is predicted value, and L is actual issued transaction amount per second, when trading volume is higher than certain threshold value, can use relative error:
S4, predicts the outcome with image conversion form display TPS transaction event threshold value.
Particularly, as shown in Figure 3, after regression model has been trained, input test collection data can carry out the prediction of TPS transaction event threshold value, such as, TPS predicted the method that changes of threshold figure displays in visual form.Show result as shown in Figure 5.
In an embodiment of the present invention, history bank backstage transaction data can be utilized, dope next such as: week age, every the arm's length transaction weight range of 5 minutes, i.e. threshold value.And by showing TPS transaction event threshold value, improve to the background service of bank and reference is provided, also suggestion can be provided to the decision-making of bank's troubleshooting methodology, that is, commercial bank can be helped to improve background service, also can offer suggestions for the fast quick-recovery of fault after bank breaks down.
According to the Forecasting Methodology of the TPS transaction event threshold value that the embodiment of the present invention proposes, by extracting TPS data and TPS data characteristics from transaction data, to add up historical trading maximal value corresponding to each time window and minimum value and to obtain TPS transaction event prediction regression model, thus prediction TPS transaction event threshold value, improve the reliability that data are extracted, realize the object of prediction TPS transaction event threshold value, not only provide reference for background service improves, and suggestion is provided to the decision-making of failture evacuation, improve the experience of user, simple and convenient.
The prognoses system of the TPS transaction event threshold value proposed according to the embodiment of the present invention is described with reference to the accompanying drawings.With reference to shown in Fig. 6, this prognoses system 100 comprises: data preprocessing module 10, statistical module 20, regression block 30 and prediction module 40.
Wherein, data preprocessing module 10 for obtaining transaction data, and extracts TPS data and TPS data characteristics from transaction data.Statistical module 20 for carrying out time series segmentation according to time window to TPS data, to add up historical trading maximal value corresponding to each time window and minimum value.Regression block 30 is for predicting regression model according to TPS data characteristics training data mining model to obtain TPS transaction event.Prediction module 40 is for predicting the historical trading maximal value that regression model and each time window are corresponding and minimum value prediction TPS transaction event threshold value according to TPS transaction event.The prognoses system of the embodiment of the present invention can not only be improved to background service and provide reference, also can provide suggestion to the decision-making of troubleshooting methodology.
Further, in one embodiment of the invention, as shown in Figure 7, the prognoses system 100 of the embodiment of the present invention also comprises: graphic user interface 50.Wherein, graphic user interface 50 is predicted the outcome for being shown TPS transaction event threshold value by image conversion mode.
Further, in one embodiment of the invention, regression block 30 is also for selecting data mining model according to type of transaction, and wherein, data mining model comprises k Neighborhood Model, BP neural network model or Random Forest model.
Further, in one embodiment of the invention, TPS data characteristics is the feature that the multiple relevant information extracted from transaction data connects each moment formed, wherein, multiple relevant information comprises one or more in currency transaction information, current date week, in the past contemporaneous information and day trade transaction amplification information.
Further, in one embodiment of the invention, from transaction data, TPS data characteristics is extracted by principal component analysis (PCA) or Method for Feature Selection.
Particularly, be predicted as example with the TPS transaction event on bank backstage and be described, it should be noted that the prognoses system 100 of the embodiment of the present invention is not limited only to a kind of application platform of bank.
In one particular embodiment of the present invention, data preprocessing module 10 is for extracting TPS data and feature thereof from bank's backstage transaction data, wherein, TPS data characteristics refers to that the multiple relevant information extracted from bank's backstage transaction data links and the feature in each moment of formation, such as, extracted the TPS data characteristics in each moment from bank's backstage transaction data by Method for Feature Selection.Statistical module 20 can be added up historical trading data, obtains maximum trading volume and minimum trading volume.Regression block 30 is for selecting k Neighborhood Model, BP neural network model or Random Forest model, and according to TPS data training k Neighborhood Model, BP neural network model or Random Forest model to obtain the TPS transaction event threshold value prediction regression model of having trained, and to TPS transaction event threshold value prediction regression model input test collection data to carry out the prediction of TPS transaction event threshold value.Graphic user interface 50 is for predicting the outcome with image conversion form display TPS transaction event threshold value.
Further, be first the preparatory stage being; Next is the operational phase of system.In the system preparatory stage, mainly need the bank TPS data analysis on backstage.First, system is for the bank's backstage trading volume data in some specific time period, extract the moment feature being such as interval with 5 minutes, then trading volume statistics is carried out, obtain the maximum trading volume of each moment history and minimum trading volume, then automatically select such as k Neighborhood Model, BP neural network model according to different types of data or carry out the training of random forest regression model, obtaining k Neighborhood Model, BP neural network model or Random Forest model after training.In the operational phase of system, user can use system to predict the TPS transaction event threshold value in 1 week, and system represents result in visual form, finds threshold value and the amplification situation of bank's backstage TPS transaction event.
In an embodiment of the present invention, the embodiment of the present invention can utilize history bank backstage transaction data (as in a week), dopes next such as: 7 days, every the arm's length transaction scope of 5 minutes, comprise maximum trading volume and minimum trading volume.And by showing TPS transaction event threshold value, improve to the background service of bank and reference is provided, also suggestion can be provided to the decision-making of bank's troubleshooting methodology, that is, commercial bank can be helped to improve background service, also can offer suggestions for the fast quick-recovery of fault after bank breaks down.
Further, with reference to shown in Fig. 7, the prognoses system 100 of the embodiment of the present invention can be divided into three large primary layers, and top layer is subscriber interface module 50; Centre is statistical module 20 and regression block 30; Bottom is data preprocessing module 10.
Wherein, subscriber interface module 50 provides the user interface of a patterned close friend mainly to the user of TPS transaction event threshold value prognoses system, carries out the prediction of TPS transaction event threshold value to facilitate user.
Statistical module 20 and regression block 30 provide whole system historical trading volume data statistics, regression model training, prediction interface, and visual TPS threshold value prediction show.
Data preprocessing module 10 mainly feature extraction, such as, comprise following relevant information: 1) current trading situation: using the transaction equal temperance of front for the same day 5 minutes, 10 minutes, 1 hour as feature; 2) week date on the same day: due to the end of month middle of the month at the beginning of the month, and between Mon-Fri and festivals or holidays at weekend, trade off curve all may have a great difference; 3) trading situation in the past: the characteristic of getting a week is as feature same period in the past; 4) the amplification situation of day trade transaction: use the difference between day trade transaction amount.These 4 relevant informations are linked into the feature in this moment.
It should be noted that, the system specific implementation of the embodiment of the present invention and the specific implementation of method part similar, in order to reduce redundancy, be not described in detail herein.
According to the prognoses system of the TPS transaction event threshold value that the embodiment of the present invention proposes, by extracting TPS data and TPS data characteristics from transaction data, to add up historical trading maximal value corresponding to each time window and minimum value and to obtain TPS transaction event prediction regression model, thus prediction TPS transaction event threshold value, improve the reliability that data are extracted, realize the object of prediction TPS transaction event threshold value, not only provide reference for background service improves, and suggestion is provided to the decision-making of failture evacuation, improve the experience of user, simple and convenient.
Describe and can be understood in process flow diagram or in this any process otherwise described or method, represent and comprise one or more for realizing the module of the code of the executable instruction of the step of specific logical function or process, fragment or part, and the scope of the preferred embodiment of the present invention comprises other realization, wherein can not according to order that is shown or that discuss, comprise according to involved function by the mode while of basic or by contrary order, carry out n-back test, this should understand by embodiments of the invention person of ordinary skill in the field.
In flow charts represent or in this logic otherwise described and/or step, such as, the sequencing list of the executable instruction for realizing logic function can be considered to, may be embodied in any computer-readable medium, for instruction execution system, device or equipment (as computer based system, comprise the system of processor or other can from instruction execution system, device or equipment instruction fetch and perform the system of instruction) use, or to use in conjunction with these instruction execution systems, device or equipment.With regard to this instructions, " computer-readable medium " can be anyly can to comprise, store, communicate, propagate or transmission procedure for instruction execution system, device or equipment or the device that uses in conjunction with these instruction execution systems, device or equipment.The example more specifically (non-exhaustive list) of computer-readable medium comprises following: the electrical connection section (electronic installation) with one or more wiring, portable computer diskette box (magnetic device), random access memory (RAM), ROM (read-only memory) (ROM), erasablely edit ROM (read-only memory) (EPROM or flash memory), fiber device, and portable optic disk ROM (read-only memory) (CDROM).In addition, computer-readable medium can be even paper or other suitable media that can print described program thereon, because can such as by carrying out optical scanning to paper or other media, then carry out editing, decipher or carry out process with other suitable methods if desired and electronically obtain described program, be then stored in computer memory.
Should be appreciated that each several part of the present invention can realize with hardware, software, firmware or their combination.In the above-described embodiment, multiple step or method can with to store in memory and the software performed by suitable instruction execution system or firmware realize.Such as, if realized with hardware, the same in another embodiment, can realize by any one in following technology well known in the art or their combination: the discrete logic with the logic gates for realizing logic function to data-signal, there is the special IC of suitable combinational logic gate circuit, programmable gate array (PGA), field programmable gate array (FPGA) etc.
Those skilled in the art are appreciated that realizing all or part of step that above-described embodiment method carries is that the hardware that can carry out instruction relevant by program completes, described program can be stored in a kind of computer-readable recording medium, this program perform time, step comprising embodiment of the method one or a combination set of.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing module, also can be that the independent physics of unit exists, also can be integrated in a module by two or more unit.Above-mentioned integrated module both can adopt the form of hardware to realize, and the form of software function module also can be adopted to realize.If described integrated module using the form of software function module realize and as independently production marketing or use time, also can be stored in a computer read/write memory medium.
The above-mentioned storage medium mentioned can be ROM (read-only memory), disk or CD etc.
In the description of this instructions, specific features, structure, material or feature that the description of reference term " embodiment ", " some embodiments ", " example ", " concrete example " or " some examples " etc. means to describe in conjunction with this embodiment or example are contained at least one embodiment of the present invention or example.In this manual, identical embodiment or example are not necessarily referred to the schematic representation of above-mentioned term.And the specific features of description, structure, material or feature can combine in an appropriate manner in any one or more embodiment or example.
Although illustrate and describe embodiments of the invention above, be understandable that, above-described embodiment is exemplary, can not be interpreted as limitation of the present invention, those of ordinary skill in the art can change above-described embodiment within the scope of the invention when not departing from principle of the present invention and aim, revising, replacing and modification.
Claims (10)
1. a Forecasting Methodology for TPS transaction event threshold value, is characterized in that, comprises the following steps:
Obtain transaction data, and from described transaction data, extract issued transaction amount TPS data per second and TPS data characteristics;
According to time window, time series segmentation is carried out to described TPS data, to add up historical trading maximal value corresponding to each time window and minimum value;
According to described TPS data characteristics training data mining model to obtain TPS transaction event prediction regression model; And
According to described TPS transaction event prediction regression model and historical trading maximal value corresponding to each time window described and minimum value prediction TPS transaction event threshold value.
2. the Forecasting Methodology of TPS transaction event threshold value according to claim 1, is characterized in that, also comprise: show TPS transaction event threshold value by image conversion mode and predict the outcome.
3. the Forecasting Methodology of TPS transaction event threshold value according to claim 1, is characterized in that, select described data mining model according to type of transaction, and wherein, described data mining model comprises k Neighborhood Model, BP neural network model or Random Forest model.
4. the Forecasting Methodology of TPS transaction event threshold value according to claim 1, it is characterized in that, described TPS data characteristics is the feature that the multiple relevant information extracted from described transaction data connects each moment formed, wherein, described multiple relevant information comprises one or more in currency transaction information, current date week, in the past contemporaneous information and day trade transaction amplification information.
5. the Forecasting Methodology of the TPS transaction event threshold value according to any one of claim 1-4, is characterized in that, extracts TPS data characteristics by principal component analysis (PCA) or Method for Feature Selection from described transaction data.
6. a prognoses system for TPS transaction event threshold value, is characterized in that, comprising:
Data preprocessing module, for obtaining transaction data, and extracts TPS data and TPS data characteristics from described transaction data;
Statistical module, for carrying out time series segmentation according to time window to described TPS data, to add up historical trading maximal value corresponding to each time window and minimum value;
Regression block, for predicting regression model according to described TPS data characteristics training data mining model to obtain TPS transaction event; And
Prediction module, for according to described TPS transaction event prediction regression model and historical trading maximal value corresponding to each time window described and minimum value prediction TPS transaction event threshold value.
7. the prognoses system of TPS transaction event threshold value according to claim 6, is characterized in that, also comprise: graphic user interface, predicts the outcome for being shown TPS transaction event threshold value by image conversion mode.
8. the prognoses system of TPS transaction event threshold value according to claim 6, it is characterized in that, described regression block is also for selecting described data mining model according to type of transaction, and wherein, described data mining model comprises k Neighborhood Model, BP neural network model or Random Forest model.
9. the prognoses system of TPS exchange hour threshold value according to claim 6, it is characterized in that, described TPS data characteristics is the feature that the multiple relevant information extracted from described transaction data connects each moment formed, wherein, described multiple relevant information comprises one or more in currency transaction information, current date week, in the past contemporaneous information and day trade transaction amplification information.
10. the prognoses system of the TPS exchange hour threshold value according to any one of claim 6-9, is characterized in that, extracts TPS data characteristics by principal component analysis (PCA) or Method for Feature Selection from described transaction data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510213263.5A CN104778622A (en) | 2015-04-29 | 2015-04-29 | Method and system for predicting TPS transaction event threshold value |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510213263.5A CN104778622A (en) | 2015-04-29 | 2015-04-29 | Method and system for predicting TPS transaction event threshold value |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104778622A true CN104778622A (en) | 2015-07-15 |
Family
ID=53620072
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510213263.5A Pending CN104778622A (en) | 2015-04-29 | 2015-04-29 | Method and system for predicting TPS transaction event threshold value |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104778622A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105844501A (en) * | 2016-05-18 | 2016-08-10 | 上海亿保健康管理有限公司 | Consumption behavior risk control system and method |
| CN106503841A (en) * | 2016-10-17 | 2017-03-15 | 东软集团股份有限公司 | The determination method and apparatus of metrics-thresholds |
| CN108334521A (en) * | 2017-01-19 | 2018-07-27 | 阿里巴巴集团控股有限公司 | A kind of database volume prediction technique and device |
| CN109146128A (en) * | 2018-06-29 | 2019-01-04 | 阿里巴巴集团控股有限公司 | Business data processing method, device and server |
| CN109213800A (en) * | 2018-07-25 | 2019-01-15 | 山东中烟工业有限责任公司 | A kind of tobacco insect pest situation forecasting system and method |
| WO2019179223A1 (en) * | 2018-03-20 | 2019-09-26 | 阿里巴巴集团控股有限公司 | Transaction volume prediction method and device |
| CN111611146A (en) * | 2020-06-18 | 2020-09-01 | 南方电网科学研究院有限责任公司 | Micro-service fault prediction method and device |
| CN111814986A (en) * | 2020-07-07 | 2020-10-23 | 上海交通大学包头材料研究院 | Dynamic network flow controller scheduling and service type distribution method and controller algorithm |
| WO2021032056A1 (en) * | 2019-08-21 | 2021-02-25 | 深圳前海微众银行股份有限公司 | Method and apparatus for processing batch tasks, computing device and storage medium |
| CN113393061A (en) * | 2021-08-17 | 2021-09-14 | 腾讯科技（深圳）有限公司 | Transaction packing method based on block chain and related device |
| TWI792101B (en) * | 2019-11-25 | 2023-02-11 | 南韓商迪艾股份有限公司 | Data Quantification Method Based on Confirmed Value and Predicted Value |
| WO2023024679A1 (en) * | 2021-08-27 | 2023-03-02 | 深圳前海微众银行股份有限公司 | Method and apparatus for predicting server capacity |
| WO2023087705A1 (en) * | 2021-11-17 | 2023-05-25 | 深圳前海微众银行股份有限公司 | Resource prediction method and apparatus |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101430660A (en) * | 2008-11-18 | 2009-05-13 | 山东浪潮齐鲁软件产业股份有限公司 | Pressure model analysis method based on TPS in software performance test |
| CN104123592A (en) * | 2014-07-15 | 2014-10-29 | 清华大学 | Method and system for predicting transaction per second (TPS) transaction events of bank background |
| US20150081492A1 (en) * | 2013-09-16 | 2015-03-19 | International Business Machines Corporation | Transactional risk daily limit update alarm |
-
2015
- 2015-04-29 CN CN201510213263.5A patent/CN104778622A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101430660A (en) * | 2008-11-18 | 2009-05-13 | 山东浪潮齐鲁软件产业股份有限公司 | Pressure model analysis method based on TPS in software performance test |
| US20150081492A1 (en) * | 2013-09-16 | 2015-03-19 | International Business Machines Corporation | Transactional risk daily limit update alarm |
| CN104123592A (en) * | 2014-07-15 | 2014-10-29 | 清华大学 | Method and system for predicting transaction per second (TPS) transaction events of bank background |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105844501A (en) * | 2016-05-18 | 2016-08-10 | 上海亿保健康管理有限公司 | Consumption behavior risk control system and method |
| CN106503841A (en) * | 2016-10-17 | 2017-03-15 | 东软集团股份有限公司 | The determination method and apparatus of metrics-thresholds |
| CN106503841B (en) * | 2016-10-17 | 2019-12-13 | 东软集团股份有限公司 | method and device for determining index threshold |
| CN108334521A (en) * | 2017-01-19 | 2018-07-27 | 阿里巴巴集团控股有限公司 | A kind of database volume prediction technique and device |
| CN108334521B (en) * | 2017-01-19 | 2022-04-19 | 阿里巴巴集团控股有限公司 | Database capacity prediction method and device |
| WO2019179223A1 (en) * | 2018-03-20 | 2019-09-26 | 阿里巴巴集团控股有限公司 | Transaction volume prediction method and device |
| TWI690865B (en) * | 2018-03-20 | 2020-04-11 | 香港商阿里巴巴集團服務有限公司 | Transaction volume prediction method and device |
| CN109146128B (en) * | 2018-06-29 | 2022-02-18 | 创新先进技术有限公司 | Service data processing method and device and server |
| CN109146128A (en) * | 2018-06-29 | 2019-01-04 | 阿里巴巴集团控股有限公司 | Business data processing method, device and server |
| CN109213800A (en) * | 2018-07-25 | 2019-01-15 | 山东中烟工业有限责任公司 | A kind of tobacco insect pest situation forecasting system and method |
| WO2021032056A1 (en) * | 2019-08-21 | 2021-02-25 | 深圳前海微众银行股份有限公司 | Method and apparatus for processing batch tasks, computing device and storage medium |
| TWI792101B (en) * | 2019-11-25 | 2023-02-11 | 南韓商迪艾股份有限公司 | Data Quantification Method Based on Confirmed Value and Predicted Value |
| CN111611146A (en) * | 2020-06-18 | 2020-09-01 | 南方电网科学研究院有限责任公司 | Micro-service fault prediction method and device |
| CN111611146B (en) * | 2020-06-18 | 2023-05-16 | 南方电网科学研究院有限责任公司 | Micro-service fault prediction method and device |
| CN111814986A (en) * | 2020-07-07 | 2020-10-23 | 上海交通大学包头材料研究院 | Dynamic network flow controller scheduling and service type distribution method and controller algorithm |
| CN111814986B (en) * | 2020-07-07 | 2024-02-20 | 上海交通大学包头材料研究院 | Dynamic network flow controller dispatching and service type distribution method |
| CN113393061A (en) * | 2021-08-17 | 2021-09-14 | 腾讯科技（深圳）有限公司 | Transaction packing method based on block chain and related device |
| WO2023024679A1 (en) * | 2021-08-27 | 2023-03-02 | 深圳前海微众银行股份有限公司 | Method and apparatus for predicting server capacity |
| WO2023087705A1 (en) * | 2021-11-17 | 2023-05-25 | 深圳前海微众银行股份有限公司 | Resource prediction method and apparatus |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104778622A (en) | Method and system for predicting TPS transaction event threshold value | |
| Sigrist et al. | Grabit: Gradient tree-boosted Tobit models for default prediction | |
| CN104123592B (en) | Bank's backstage TPS transaction events trend forecasting method and system | |
| KR102044205B1 (en) | Target information prediction system using big data and machine learning and method thereof | |
| CN110866819A (en) | Automatic credit scoring card generation method based on meta-learning | |
| CN110738564A (en) | Post-loan risk assessment method and device and storage medium | |
| KR102330423B1 (en) | Online default forecasting system using image recognition deep learning algorithm | |
| CN114048436A (en) | Construction method and construction device for forecasting enterprise financial data model | |
| CN113537807B (en) | Intelligent wind control method and equipment for enterprises | |
| CN112700324A (en) | User loan default prediction method based on combination of Catboost and restricted Boltzmann machine | |
| CN111738504A (en) | Enterprise financial index fund amount prediction method and device, equipment and storage medium | |
| Scherer et al. | On the practical art of state definitions for Markov decision process construction | |
| Neal et al. | Forecasting the world city network | |
| Lopes et al. | Predicting recovery of credit operations on a brazilian bank | |
| Camelia et al. | A Computational Grey Based Model for Companies Risk Forecasting. | |
| Wang et al. | A Transformer-based multi-entity load forecasting method for integrated energy systems | |
| Gleue et al. | Decision support for the automotive industry: Forecasting residual values using artificial neural networks | |
| KR102499182B1 (en) | Loan regular auditing system using artificia intellicence | |
| US20210350464A1 (en) | Quantitative customer analysis system and method | |
| CN114612239A (en) | Stock public opinion monitoring and wind control system based on algorithm, big data and artificial intelligence | |
| CN112785414B (en) | Credit risk prediction method based on knowledge graph and ontology inference engine | |
| US20230118645A1 (en) | System for debt collection | |
| Gharbi | Coarse-Grained Process Diagnostics: A Method Combining Process Mining and Time Series Analysis | |
| Ozgulbas et al. | Application of data mining method for financial profiling | |
| Severeijns | Challenging LGD models with Machine Learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150715 |