CN113516559A - Fund risk determining method and device - Google Patents

Fund risk determining method and device Download PDF

Info

Publication number
CN113516559A
CN113516559A CN202110515484.3A CN202110515484A CN113516559A CN 113516559 A CN113516559 A CN 113516559A CN 202110515484 A CN202110515484 A CN 202110515484A CN 113516559 A CN113516559 A CN 113516559A
Authority
CN
China
Prior art keywords
model
fund
training
machine learning
data
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
Application number
CN202110515484.3A
Other languages
Chinese (zh)
Inventor
卢业
郭锐鹏
林露蕃
谢超
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Industrial and Commercial Bank of China Ltd ICBC
Original Assignee
Industrial and Commercial Bank of China Ltd ICBC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Industrial and Commercial Bank of China Ltd ICBC filed Critical Industrial and Commercial Bank of China Ltd ICBC
Priority to CN202110515484.3A priority Critical patent/CN113516559A/en
Publication of CN113516559A publication Critical patent/CN113516559A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/06Asset management; Financial planning or analysis
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/21Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G06F18/2135Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/04Architecture, e.g. interconnection topology
    • G06N3/045Combinations of networks
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/04Architecture, e.g. interconnection topology
    • G06N3/047Probabilistic or stochastic networks
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/08Learning methods

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Mathematical Physics (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Business, Economics & Management (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Software Systems (AREA)
  • Biophysics (AREA)
  • Development Economics (AREA)
  • Finance (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Accounting & Taxation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Probability & Statistics with Applications (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Game Theory and Decision Science (AREA)
  • Human Resources & Organizations (AREA)
  • Operations Research (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Technology Law (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides a fund risk determination method and a fund risk determination device, which relate to the field of finance, and the method comprises the following steps: obtaining basic metal data; respectively carrying out model training on a pre-established exponential conditional variance model and a machine learning model by utilizing the basic metal data to determine an exponential conditional variance prediction model and a machine learning prediction model; and generating a fund risk determination result by using the index condition variance prediction model and the machine learning prediction model determined by the basic metal data and training. The invention utilizes the prediction results of the two models to generate a fund risk determination result, provides a fluctuation rate prediction model integrating two model algorithms, and compared with a scheme adopting one prediction model, the overall performance of the invention is superior to that of other comparison models, and the prediction accuracy is better.

Description

Fund risk determining method and device
Technical Field
The invention relates to a data processing technology, in particular to a fund risk determination method and device.
Background
At present, the prediction methods of the internet fund product risk prediction model are roughly divided into two types in the relevant fund prediction literature: (1) a statistical method and (2) an artificial intelligence method. The statistical method comprises autoregression, moving average and autoregression condition variance.
The artificial intelligence method is mainly a neural network method, and is different from the traditional statistical model, the neural network is a data-driven nonparametric model, and the neural network enables data to speak by itself, so that the artificial intelligence method has a greater potential in describing the dynamics of the fund. But neural networks also have a number of disadvantages, above all, it is easily involved in over-learning, which is based on the empirical risk minimization principle of neural networks. The over-learning problem reduces the prediction accuracy of the neural network, and at the same time, the solution of the neural network is likely to be locally optimal rather than globally optimal, which will also affect the prediction accuracy of the model.
The GARCH model based on traditional statistics is limited by a linear structure of the model, and the nonlinear components in time series data cannot be effectively predicted, so that the prediction precision in practical application needs to be improved, and the prediction effect is often unsatisfactory.
Disclosure of Invention
In order to overcome a defect in the existing fund risk prediction, the invention provides a fund risk determination method, which comprises the following steps:
obtaining basic metal data;
respectively carrying out model training on a pre-established exponential conditional variance model and a machine learning model by utilizing the basic metal data to determine an exponential conditional variance prediction model and a machine learning prediction model;
and generating a fund risk determination result by using the index condition variance prediction model and the machine learning prediction model determined by the basic metal data and training.
In an embodiment of the present invention, the basic metallic data includes: fund yield data, fund scale data, stock market closing price data, generalized currency supply amount data and national debt index data.
In an embodiment of the present invention, the exponential conditional variance model includes: EGARCH model;
the machine learning model comprises: EFKOS-GELM model.
In the embodiment of the present invention, the determining the index conditional variance prediction model and the machine learning prediction model by performing model training on the pre-established index conditional variance model and the machine learning model using the basic metal data includes:
performing initialization training on a pre-established EFKOS-GELM model by using the attribute data to determine an initialization training model;
and performing sequential learning training on the initialization training model by using the attribute data to determine an EFKOS-GELM prediction model.
In an embodiment of the present invention, the generating a fund risk determination result by using the exponential conditional variance prediction model and the machine learning prediction model determined by the basic metal data and training includes:
generating a fund fluctuation rate predicted value of an exponential conditional variance model by using the basic metal data and the exponential conditional variance prediction model determined by training;
generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training;
and generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential condition variance model and the fund fluctuation rate predicted value of the machine learning model.
In an embodiment of the present invention, the generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training includes:
performing principal component analysis processing on the basic metal data to generate prediction data;
and generating a fund fluctuation rate predicted value of the machine learning model by using the generated predicted data and the machine learning predicted model determined by training.
In an embodiment of the present invention, the generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential conditional variance model and the fund fluctuation rate predicted value of the machine learning model includes:
and mixing the fund fluctuation rate predicted value of the exponential conditional variance model and the fund fluctuation rate predicted value of the machine learning model according to a preset weighted mixing proportion to generate a fund risk determination result.
Meanwhile, the invention also provides a fund risk determination device, which comprises:
the data acquisition module is used for acquiring basic metal data;
the model training module is used for respectively carrying out model training on a pre-established exponential conditional variance model and a machine learning model by utilizing the basic metal data to determine an exponential conditional variance prediction model and a machine learning prediction model;
and the risk determination module is used for generating a fund risk determination result by utilizing the basic metal data and the index condition variance prediction model and the machine learning prediction model determined by training.
In an embodiment of the present invention, the model training module includes:
the initialization unit is used for performing initialization training on a pre-established EFKOS-GELM model by utilizing the attribute data to determine an initialization training model;
and the EFKOS-GELM prediction model determining unit is used for performing sequential learning training on the initialization training model by using the attribute data to determine the EFKOS-GELM prediction model.
In an embodiment of the present invention, the risk determining module includes:
the first prediction fluctuation rate determining unit is used for generating a fund fluctuation rate prediction value of an exponential conditional variance model by using the basic metal data and the exponential conditional variance prediction model determined by training;
the second prediction fluctuation rate determining unit is used for generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training;
and the mixing unit is used for generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential condition variance model and the fund fluctuation rate predicted value of the machine learning model.
Meanwhile, the invention also provides computer equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor realizes the method when executing the computer program.
Meanwhile, the invention also provides a computer readable storage medium, and a computer program for executing the method is stored in the computer readable storage medium.
The fund risk determination method and the fund risk determination device respectively perform model training on a pre-established index conditional variance model and a pre-established machine learning model by utilizing the basic metal data to determine an index conditional variance prediction model and a machine learning prediction model, and generate a fund risk determination result by utilizing the basic metal data and the index conditional variance prediction model and the machine learning prediction model determined by training. Compared with the traditional GARCH model and the OS-GELM model, the invention has better overall performance than other comparison models and better prediction accuracy.
In order to make the aforementioned and other objects, features and advantages of the invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of a fund risk determination method provided by the present invention;
FIG. 2 is a block diagram of a fund risk prediction system based on intelligent search according to an embodiment of the present invention;
FIG. 3 is a block diagram of a fund prediction model training platform according to an embodiment of the present invention;
FIG. 4 is a block diagram of a hybrid model training structure based on EGARCH and EFKOS-GELM of the model training unit in the embodiment of the present invention;
FIG. 5 is a flowchart illustrating the training of the EFKOS-GELM model according to an embodiment of the present invention;
FIG. 6 is a flow chart of the construction of a hybrid model based on EGARCH and EFKOS-GELM according to the present invention;
FIG. 7 is a block diagram of an intelligent retrieval system in accordance with an embodiment of the present invention;
FIG. 8 is a block diagram of a fund risk determination apparatus provided by an embodiment of the present invention;
fig. 9 is a schematic diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to overcome the limitation that the prior internet fund products mainly adopt a qualitative method, a new internet fund product market risk prediction model is constructed, the defects of time consumption and high cost in artificial neural network training are overcome, the risk of the fund products is predicted by constructing a fluctuation rate prediction model mixing a GARCH model and a machine learning model, and a risk prevention rating is formed through an intelligent retrieval system so as to make auxiliary reference for users to purchase the fund products.
Description of terms:
kalman filtering theory: in order to be able to solve the problem of processing real-time signals, the filtering theory was further improved and optimized by the well-known scientist Kalman in the 60's of the 20 th century. On the basis of this, he proposes the well-known kalman filtering theory. The Kalman filtering algorithm is an optimal estimation algorithm designed by utilizing a state space model, and requires that a state equation and a measurement equation of a system of the Kalman filtering algorithm must be established, and then recursive filtering is performed on the basis, so that an optimal estimation result of a state can be obtained in real time.
Coefficient of variation method: the coefficient of variation method is an objective weighting method for calculating the weight of an index by directly using information included in each index. The basic idea of this method is: the weight of each evaluation index is determined by the degree of difference between each average index.
Principle of Principal Component Analysis (PCA): principal component analysis is a statistical method by which the dimensionality of data can be reduced while preserving the original data. On the premise of losing little information, a group of variables possibly having correlation are processed in an orthogonal transformation mode to obtain a group of linear uncorrelated variables, and the number of the variables after conversion is usually less than that before conversion, that is, the dimension reduction processing is performed on a high-dimensional variable space. The core of the method is that a few characteristic variables are extracted according to a certain standard to be used as main components to be analyzed and calculated, and the extracted standard mainly has characteristic roots larger than 1 or the cumulative contribution rate of characteristic values exceeding a certain proportion (such as 85%).
In order to overcome the defects of time consumption and high cost in artificial neural network training in fund risk prediction in the prior art, the invention provides a fund risk determination method, as shown in fig. 1, the fund risk determination method provided by the invention comprises the following steps:
step S101, obtaining basic metal data;
step S102, model training is respectively carried out on a pre-established exponential conditional variance model and a machine learning model by utilizing the basic metal data to determine an exponential conditional variance prediction model and a machine learning prediction model;
and step S103, generating a fund risk determination result by using the basic metal data, the index conditional variance prediction model determined by training and the machine learning prediction model.
The fund risk determination method provided by the invention utilizes an exponential condition variance prediction model and a machine learning prediction model determined by training, namely, the fund risk determination result is generated by utilizing the prediction results of the two models, and a fluctuation rate prediction model integrating two model algorithms is provided.
In the embodiment of the present invention, the basic metallic data includes: fund yield data, fund scale data, stock market closing price data, generalized currency supply amount data and national debt index data.
In one embodiment of the present invention, the basic metallic data includes but is not limited to: fund yield, fund size, Shanghai 300 index closing price, Shanghai comprehensive index closing price, deep component index closing price, generalized currency supply M2, Chinese debt national debt total index and Chinese debt inter-bank national debt index.
Further, in the embodiment of the present invention, the exponential conditional variance model includes: EGARCH model, i.e. exponential conditional heteroscedastic model.
The machine learning model includes: the EFKOS-GELM model is an OS-GELM model integrating a forgetting mechanism and Kalman filtering. According to the model, a forgetting mechanism and a Kalman filtering theory are added into the OS-GELM model, the problem of timeliness of time series data and the problem of multiple collinearity possibly existing in the OS-GELM model when the output weight is updated are solved, and the stability and the accuracy of the OS-GELM model are further improved by using an integration idea.
The fluctuation rate prediction model integrated based on the EGARCH model and the EFKOS-GELM algorithm provided by the embodiment of the invention considers the possible correlation among input variables, and compared with the traditional GARCH model and the OS-GELM model, the overall performance of the invention is better than that of other comparison models, and the prediction accuracy is better. A single prediction model, whether a linear model or a nonlinear model, can not well process complex practical problems, so on the basis of an EFKOS-GELM model, the EGARCH model and the EFKOS-GELM model are linearly mixed, and simultaneously, principal component analysis is carried out on input data, so that the prediction error is minimum, and the prediction effect is optimal.
In the embodiment of the present invention, the determining the index conditional variance prediction model and the machine learning prediction model by performing model training on the pre-established index conditional variance model and the machine learning model using the basic metal data includes:
performing initialization training on a pre-established EFKOS-GELM model by using the attribute data to determine an initialization training model;
and performing sequential learning training on the initialization training model by using the attribute data to determine an EFKOS-GELM prediction model.
In an embodiment of the present invention, the generating a fund risk determination result by using the exponential conditional variance prediction model and the machine learning prediction model determined by the basic metal data and training includes:
generating a fund fluctuation rate predicted value of an exponential conditional variance model by using the basic metal data and the exponential conditional variance prediction model determined by training;
generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training;
and generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential condition variance model and the fund fluctuation rate predicted value of the machine learning model.
In an embodiment of the present invention, the generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training includes:
performing principal component analysis processing on the basic metal data to generate prediction data;
and generating a fund fluctuation rate predicted value of the machine learning model by using the generated predicted data and the machine learning predicted model determined by training.
And processing the basic metal data by using a principal component analysis method. Considering that the input data may have a certain correlation among the indexes due to a large number of dimensions, and redundant information exists, this problem may bring a certain noise and negatively affect the result of prediction. Therefore, it is necessary to perform a PCA (principal component analysis) method to perform a dimensionality reduction process on the data, extract a small amount of principal components to replace the original input data, reduce redundant information, and further improve the prediction accuracy.
In an embodiment of the present invention, the generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential conditional variance model and the fund fluctuation rate predicted value of the machine learning model includes:
and mixing the fund fluctuation rate predicted value of the exponential conditional variance model and the fund fluctuation rate predicted value of the machine learning model according to a preset weighted mixing proportion to generate a fund risk determination result.
Namely, in the embodiment of the invention, the EGARCH model is applied to predict the fluctuation rate of the fund yield. An artificial neural network (EFKOS-GELM model) is applied to predict the volatility of the fund yield. And (3) finishing the training of the extracted main component by applying an OS-GELM model (EFKOS-GELM model) integrating a forgetting mechanism and Kalman filtering, and predicting the fluctuation rate by applying the obtained optimal model after obtaining the optimal model so as to obtain a fund fluctuation rate predicted value.
And mixing the EGARCH model prediction result and the EFKOS-GELM model prediction result to obtain a fluctuation rate prediction result of the mixed model.
In this embodiment, the fluctuation rate prediction result of the EGARCH model and the fluctuation rate prediction result of the EFKOS-GELM model are mixed by a weighted average method, so as to obtain the final prediction result of the fluctuation rate of the hybrid model constructed in the embodiment of the present invention.
The specific mixing manner of this example is as follows:
Y=α×Y1+(1-α)×Y2
wherein Y represents the fluctuation rate prediction result of the mixed model, and Y1Denotes EFKOS-Volatility prediction of the GELM model, Y2The method is characterized in that the method represents the fluctuation rate prediction result of the EGARCH model, and alpha represents the mixing ratio of the fluctuation rate prediction results of the two models. The mixing ratio alpha is determined by the coefficient of variation method.
The fund risk determining method provided by the embodiment of the invention overcomes the limitation that the prior internet fund products mainly adopt a qualitative method, constructs a new internet fund product market risk prediction model, overcomes the defects of time consumption and high cost in artificial neural network training, constructs a fluctuation rate prediction model mixing a GARCH model and a machine learning model to predict the risk of the fund product, and forms risk prevention rating through an intelligent retrieval system so as to make auxiliary reference for users to purchase fund products.
Specifically, in an embodiment of the invention, the fund risk prediction value is determined according to the trained model, and the specific fund risk value and the recommended fund product result are displayed on the front-end page in a visual display mode according to the tolerance threshold set by the user, so that fund risk prediction reminding is provided for the user.
Fig. 2 is a block diagram of a fund risk prediction system based on intelligent retrieval according to an embodiment of the present invention, and as shown in fig. 2, the fund risk prediction system based on intelligent retrieval according to the embodiment includes:
the fund prediction model training platform comprises a fund prediction model training platform 1 and an intelligent retrieval system 2. The fund prediction model training platform 1 can be connected with the intelligent retrieval system 2 through a mobile or wired network; specifically, the method comprises the following steps:
and the fund prediction model training platform 1 is responsible for collecting sample data of various funds, performing mixed model training and calculating the risk metric value of the fund product.
The fund prediction model training platform 1 is used for collecting basic metal data, and respectively carrying out model training on a pre-established index condition variance model and a machine learning model by utilizing the basic metal data to determine the index condition variance prediction model and the machine learning prediction model; and generating a fund risk determination result by using the index condition variance prediction model and the machine learning prediction model determined by the basic metal data and training.
And the intelligent retrieval system 2 is used for receiving the risk metric values of the fund products, feeding back the risk values set under different threshold values to the user in time, and searching and recommending the fund products to the user according to the threshold values.
Fig. 3 is a block diagram showing a structure of a fund prediction model training platform 1 in the fund risk prediction system based on intelligent retrieval, and as shown in fig. 3, the fund prediction model training platform 1 includes a data acquisition unit 11, a data cleaning unit 12, a model training unit 13, and a data storage unit 14; wherein:
the data acquisition unit 11 is responsible for utilizing the crawler technology and the scanning monitoring technology to acquire the data required by the model training unit 13 relatively comprehensively, and in this embodiment, the acquired training data includes: fund yield, fund size, Shanghai 300 index closing price, Shanghai comprehensive index closing price, deep component index closing price, generalized currency supply M2, Chinese debt national debt total index and Chinese debt inter-bank national debt index.
And the data cleaning unit 12 is responsible for cleaning the data acquired by the crawler based on the metallic set, adjusting the data with inconsistent data formats, and removing the data with empty attribute values.
The model training unit 13 is used for performing model training based on EGARCH and EFKOS-GELM, details are shown in FIG. 4, and integrates an OS-GELM model (EFKOS-GELM model) training module 131 with a forgetting mechanism and Kalman filtering and a GARCH model (EGARCH) training module 132 for overcoming financial time series, and is responsible for predicting and analyzing risk of fund products; wherein:
the OS-GELM model (EFKOS-GELM model) training module 131 with the forgetting mechanism and the Kalman filtering is integrated and is responsible for adding the forgetting mechanism and the Kalman filtering theory to a generalized autoregressive conditional variance (GARCH) process and an online sequential extreme learning machine (OS-ELM) combination model (OS-GELM) to solve the timeliness problem of time series data, the OS-GELM model can have the problem of multiple collinearity when updating output weights, and the stability and the accuracy of the OS-GELM model are further improved by using an integration idea. The specific flow is shown in fig. 5, which is a training process of the OS-GELM model provided in this embodiment.
The GARCH model (EGARCH) training module 132, when overcoming financial time series, is responsible for predicting the volatility of the fund rate of return. And after the EGARCH model is subjected to parameter estimation to obtain the optimal parameters, the obtained optimal model is applied to predict the fluctuation rate, and the final prediction value of the fund fluctuation rate can be obtained.
The general form of the GARCH model (EGARCH) in this example appears to be:
Figure BDA0003061787120000091
et~N(0,ht)
Figure BDA0003061787120000092
wherein e represents revenue or revenue residual;
h is the conditional variance;
alpha is the current income;
Figure BDA0003061787120000093
a q-phase lag value representing a conditional variance;
Figure BDA0003061787120000094
expressing a conditional mean equation to obtain a residual error squared term to obtain a p-phase lag value;
et~N(0,ht) The mean value of e obedience is 0, and the variance is h to obtain positive-space distribution;
gamma denotes the leverage effect present in the financial market.
And the data storage unit 14 is responsible for storing the data acquired by the data acquisition unit 11 and the data analyzed and processed by the model training unit 13.
Specifically, the model training flow of the OS-GELM model (EFKOS-GELM model) training module 31 integrated with the forgetting mechanism and the kalman filter in the embodiment is shown in fig. 5, and specifically includes the following steps:
an initialization stage:
step 1: inputting all training samples; responsible for training the data set Z ═ { Z ═ Z1t,...,zit,...,zNtTime series z of | T ═ 1., T }itT is 1, …, T, i is 1, …, N, if the number of time series data used for calculating the standard deviation is s1Then the time series can be calculated using the following equation
Figure BDA0003061787120000101
Standard deviation of (2)
Figure BDA0003061787120000102
Figure BDA0003061787120000103
Wherein,
Figure BDA0003061787120000104
is a time sequence
Figure BDA0003061787120000105
Is measured. Thus, a set of standard deviations for the training data set Z can be obtained as
Figure BDA0003061787120000106
The standard deviation data set is then divided into a training data set and a test data set and initialized.
Step 2: inputting a training parameter s2
Figure BDA0003061787120000107
p; the validity period of the data block is s2Assuming an initial valid training data set at the kth unit time as
Figure BDA0003061787120000108
The number of the contained data is larger than that of the hidden layer nodes. And let r equal to 1, given
Figure BDA0003061787120000109
And p value.
Wherein r represents the r-th FKOS-GELM model;
k represents the kth time unit; j denotes a jth block data block;
p denotes that the EFKOS-GELM has p FKOS-GELM models with the same number of hidden layer nodes and the same activation function;
Figure BDA00030617871200001010
indicating the number of hidden nodes.
And step 3: randomly setting training parameters
Figure BDA00030617871200001011
And 4, step 4: inputting an initial training sample; order to
Figure BDA00030617871200001012
RkIs an identity matrix, A and QkIs a constant matrix, and RkA and QkAnd the method is kept unchanged all the time in the process of iterative calculation.
Wherein,
Figure BDA0003061787120000111
input parameters representing hidden nodes;
Figure BDA0003061787120000112
a hidden node output matrix representing the k0 time;
Figure BDA0003061787120000113
state vector representing time k 0;
Rkis an identity matrix;
a and QkIs a constant matrix;
and 5: calculating an initial iteration variable Pk0、βk0
Step 6: let r be r +1, if r ≦ p, return to step 3.
Then, sequential learning is carried out:
step 1: when the (k +1) th training data block enters the neural network, the (k-s) th training data block2+1) training data blocks are discarded, and the valid data set is now
Figure BDA0003061787120000114
Let r be 1.
Step 2: and (3) training an OS-GELM model (namely FKOS-GELM) of a forgetting mechanism and a Kalman filtering theory, and calculating corresponding variable values.
And step 3: let r be r +1, if r ≦ p, return to step 1.
And 4, step 4: and when a new training data block enters the neural network, making k equal to k +1 and r equal to 1, and returning to the step 1.
And 5: for the input data in the test data set, calculating a hidden node output matrix HrThen, the output weight vector beta calculated before is combinedrUsing the formula
Figure BDA0003061787120000115
A final prediction value can be calculated, where βrIs the output weight matrix of the r-th FKOS-GELM model.
In the embodiment of the invention, in the training of an on-line extreme learning machine (OS-ELM), the learning process of the algorithm is mainly divided into two parts:
the first part is an initialization phase, i.e. the initial output weight β of the calculation is obtained from a small number of samples0
The second part is an online learning part, and the final output weight can be obtained by iteration through the following two formulas
Figure BDA0003061787120000116
Then can getBy using the formula of Extreme Learning Machine (ELM)
Figure BDA0003061787120000117
And solving the final predicted value.
Figure BDA0003061787120000121
Figure BDA0003061787120000122
Wherein,
Figure BDA0003061787120000123
it is shown that the activation function is,
Figure BDA0003061787120000124
representing the connection weight vector between the input node and the jth hidden node,
Figure BDA0003061787120000125
indicating the offset showing the jth hidden node.
The FOS-ELM algorithm is equivalent to consider the timeliness of data on the basis of the OS-ELM algorithm, and obsolete data is discarded. Assuming that the validity period of each data block is s, the FOS-ELM algorithm trains the OS-ELM network with many sliding windows with a length s of the song window, and when the (k +1) th data block enters the neural network, we can train the OS-ELM network by applying the following equation:
Figure BDA0003061787120000126
Figure BDA0003061787120000127
the EFKOS-GELM model is a model of p FKOS-GELM models with the same number of hidden nodes and the same activation function at any unit time, then at the k +1 unit time, the output weight of the r-th FKOS-GELM model can be expressed as:
Figure BDA0003061787120000128
after training is finished, the final output weight of the r < th > FKOS-GELM model can be obtained as betarThen, combining equation (3), the final predicted value L can be obtained as:
Figure BDA0003061787120000129
Hrand (4) inputting the data to be predicted into the hidden layer node output matrix calculated by the r-th FKOS-GELM model.
Fig. 6 is a flowchart of the mixed model construction based on EGARCH and EFKOS-GELM provided in the embodiment of the present invention, and the specific steps are as follows:
step 1: and selecting the optimal GARCH model as an EGARCH model.
Step 2: the EGARCH model is applied to predict the volatility of the fund yield. And after the EGARCH model is subjected to parameter estimation to obtain the optimal parameters, the obtained optimal model is applied to predict the fluctuation rate, and the final prediction value of the fund fluctuation rate can be obtained.
And step 3: and processing the input data by applying a principal component analysis method. Considering that the input data may have a certain correlation among the indexes due to a large number of dimensions, and redundant information exists, this problem may bring a certain noise and negatively affect the result of prediction. Therefore, it is necessary to perform dimensionality reduction on data by using a PCA method, extract a small amount of principal components to replace the original input data, reduce redundant information, and further improve the prediction accuracy.
And 4, step 4: an artificial neural network (EFKOS-GELM model) is applied to predict the volatility of the fund yield. And (3) finishing the training of the extracted main component by applying an OS-GELM model (EFKOS-GELM model) training module 31 integrating a forgetting mechanism and Kalman filtering, and after obtaining an optimal model, predicting the fluctuation rate by applying the obtained optimal model to obtain a final fund fluctuation rate predicted value.
And 5: and mixing the EGARCH model prediction result and the EFKOS-GELM model prediction result to obtain a fluctuation rate prediction result of the mixed model. And mixing the fluctuation rate prediction result of the EGARCH model and the fluctuation rate prediction result of the EFKOS-GELM model by adopting a weighted average method to obtain a final prediction result of the fluctuation rate of the mixed model constructed by the patent. The specific mixing method is as follows:
Y=α×Y1+(1-α)×Y2
wherein Y represents the fluctuation rate prediction result of the mixed model, and Y1Showing the volatility prediction result of the EFKOS-GELM model, Y2The method is characterized in that the method represents the fluctuation rate prediction result of the EGARCH model, and alpha represents the mixing ratio of the fluctuation rate prediction results of the two models. The mixing ratio alpha is determined by the coefficient of variation method.
Fig. 7 is a structural diagram of the intelligent retrieval system 2 in the embodiment of the present invention, and the intelligent retrieval system 2 includes: fund risk value estimation module 21, entity information retrieval unit 22, wherein:
and the fund risk value estimation module 21 is responsible for acquiring the trained fund risk value from the data storage unit 14 in the fund prediction model training platform 1.
And the entity information retrieval unit 22 and the background end are responsible for acquiring the fund risk value from the fund risk value estimation module 21, and displaying the specific fund risk value and the recommended fund product result on a front-end page in a visual display mode according to a tolerance threshold set by a user.
The embodiment provides a fluctuation rate prediction model integrated based on an EGARCH model and an EFKOS-GELM algorithm, and simultaneously considers the possible correlation between input variables, so that the input data of the model is subjected to dimensionality reduction by using a principal component analysis method.
Considering that a single prediction model, whether a linear model or a nonlinear model, may not well handle complex practical problems, on the basis of the EFKOS-GELM model, the invention linearly mixes the EGARCH model and the EFKOS-GELM model, and simultaneously performs principal component analysis on input data, so that the prediction error is minimized, and the prediction effect is optimal.
Meanwhile, the present invention also provides a fund risk determination apparatus, as shown in fig. 8, the apparatus comprising:
a data obtaining module 801, configured to obtain basic metallic data;
the model training module 802 is configured to perform model training on a pre-established exponential conditional variance model and a pre-established machine learning model respectively by using the basic metal data to determine an exponential conditional variance prediction model and a machine learning prediction model;
and a risk determining module 803, configured to generate a fund risk determination result by using the basic metal data and the training-determined exponential conditional variance prediction model and the machine learning prediction model.
In an embodiment of the present invention, the model training module includes:
the initialization unit is used for performing initialization training on a pre-established EFKOS-GELM model by utilizing the attribute data to determine an initialization training model;
and the EFKOS-GELM prediction model determining unit is used for performing sequential learning training on the initialization training model by using the attribute data to determine the EFKOS-GELM prediction model.
In an embodiment of the present invention, the risk determining module includes:
the first prediction fluctuation rate determining unit is used for generating a fund fluctuation rate prediction value of an exponential conditional variance model by using the basic metal data and the exponential conditional variance prediction model determined by training;
the second prediction fluctuation rate determining unit is used for generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training;
and the mixing unit is used for generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential condition variance model and the fund fluctuation rate predicted value of the machine learning model.
The fund risk determination device provided by the invention utilizes the basic metal data to respectively carry out model training on the pre-established exponential conditional variance model and the machine learning model to determine the exponential conditional variance prediction model and the machine learning prediction model, and utilizes the basic metal data and the exponential conditional variance prediction model and the machine learning prediction model determined by training to generate the fund risk determination result. The method utilizes the prediction results of the two models to generate a fund risk determination result, provides a fluctuation rate prediction model integrating two model algorithms, and has better overall performance and better prediction accuracy compared with a scheme adopting one prediction model.
For those skilled in the art, the implementation of the fund risk determination apparatus provided by the present invention can be clearly understood through the foregoing description of the embodiments, and details are not described herein again.
It should be noted that the fund risk determination method and apparatus disclosed in the present invention may be used in the financial field or other fields, and the fund risk determination method and apparatus disclosed in the present invention may be used in predicting fund risk in the financial field, and may also be used in predicting risk of financial products in any fields other than the financial field.
The present embodiment also provides an electronic device, which may be a desktop computer, a tablet computer, a mobile terminal, and the like, but is not limited thereto. In this embodiment, the electronic device may refer to the embodiments of the method and the apparatus, and the contents thereof are incorporated herein, and repeated descriptions are omitted.
Fig. 9 is a schematic block diagram of a system configuration of an electronic apparatus 600 according to an embodiment of the present invention. As shown in fig. 9, the electronic device 600 may include a central processor 100 and a memory 140; the memory 140 is coupled to the central processor 100. Notably, this diagram is exemplary; other types of structures may also be used in addition to or in place of the structure to implement telecommunications or other functions.
In one embodiment, the fund risk determination function may be integrated into the central processor 100. The central processor 100 may be configured to control as follows:
obtaining basic metal data;
respectively carrying out model training on a pre-established exponential conditional variance model and a machine learning model by utilizing the basic metal data to determine an exponential conditional variance prediction model and a machine learning prediction model;
and generating a fund risk determination result by using the index condition variance prediction model and the machine learning prediction model determined by the basic metal data and training.
In an embodiment of the present invention, the basic metallic data includes: fund yield data, fund scale data, stock market closing price data, generalized currency supply amount data and national debt index data.
In an embodiment of the present invention, the exponential conditional variance model includes: EGARCH model;
the machine learning model comprises: EFKOS-GELM model.
In the embodiment of the present invention, the determining the index conditional variance prediction model and the machine learning prediction model by performing model training on the pre-established index conditional variance model and the machine learning model using the basic metal data includes:
performing initialization training on a pre-established EFKOS-GELM model by using the attribute data to determine an initialization training model;
and performing sequential learning training on the initialization training model by using the attribute data to determine an EFKOS-GELM prediction model.
In an embodiment of the present invention, the generating a fund risk determination result by using the exponential conditional variance prediction model and the machine learning prediction model determined by the basic metal data and training includes:
generating a fund fluctuation rate predicted value of an exponential conditional variance model by using the basic metal data and the exponential conditional variance prediction model determined by training;
generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training;
and generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential condition variance model and the fund fluctuation rate predicted value of the machine learning model.
In an embodiment of the present invention, the generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training includes:
performing principal component analysis processing on the basic metal data to generate prediction data;
and generating a fund fluctuation rate predicted value of the machine learning model by using the generated predicted data and the machine learning predicted model determined by training.
In an embodiment of the present invention, the generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential conditional variance model and the fund fluctuation rate predicted value of the machine learning model includes:
and mixing the fund fluctuation rate predicted value of the exponential conditional variance model and the fund fluctuation rate predicted value of the machine learning model according to a preset weighted mixing proportion to generate a fund risk determination result.
In another embodiment, the fund risk determination means may be configured separately from the central processor 100, for example, the fund risk determination means may be configured as a chip connected to the central processor 100, and the fund risk determination function is realized by the control of the central processor.
As shown in fig. 9, the electronic device 600 may further include: communication module 110, input unit 120, audio processing unit 130, display 160, power supply 170. It is noted that the electronic device 600 does not necessarily include all of the components shown in FIG. 9; furthermore, the electronic device 600 may also comprise components not shown in fig. 9, which may be referred to in the prior art.
As shown in fig. 9, the central processor 100, sometimes referred to as a controller or operational control, may include a microprocessor or other processor device and/or logic device, the central processor 100 receiving input and controlling the operation of the various components of the electronic device 600.
The memory 140 may be, for example, one or more of a buffer, a flash memory, a hard drive, a removable media, a volatile memory, a non-volatile memory, or other suitable device. The information relating to the failure may be stored, and a program for executing the information may be stored. And the central processing unit 100 may execute the program stored in the memory 140 to realize information storage or processing, etc.
The input unit 120 provides input to the cpu 100. The input unit 120 is, for example, a key or a touch input device. The power supply 170 is used to provide power to the electronic device 600. The display 160 is used to display an object to be displayed, such as an image or a character. The display may be, for example, an LCD display, but is not limited thereto.
The memory 140 may be a solid state memory such as Read Only Memory (ROM), Random Access Memory (RAM), a SIM card, or the like. There may also be a memory that holds information even when power is off, can be selectively erased, and is provided with more data, an example of which is sometimes called an EPROM or the like. The memory 140 may also be some other type of device. Memory 140 includes buffer memory 141 (sometimes referred to as a buffer). The memory 140 may include an application/function storage section 142, and the application/function storage section 142 is used to store application programs and function programs or a flow for executing the operation of the electronic device 600 by the central processing unit 100.
The memory 140 may also include a data store 143, the data store 143 for storing data, such as contacts, digital data, pictures, sounds, and/or any other data used by the electronic device. The driver storage portion 144 of the memory 140 may include various drivers of the electronic device for communication functions and/or for performing other functions of the electronic device (e.g., messaging application, address book application, etc.).
The communication module 110 is a transmitter/receiver 110 that transmits and receives signals via an antenna 111. The communication module (transmitter/receiver) 110 is coupled to the central processor 100 to provide an input signal and receive an output signal, which may be the same as in the case of a conventional mobile communication terminal.
Based on different communication technologies, a plurality of communication modules 110, such as a cellular network module, a bluetooth module, and/or a wireless local area network module, may be provided in the same electronic device. The communication module (transmitter/receiver) 110 is also coupled to a speaker 131 and a microphone 132 via an audio processor 130 to provide audio output via the speaker 131 and receive audio input from the microphone 132 to implement general telecommunications functions. Audio processor 130 may include any suitable buffers, decoders, amplifiers and so forth. In addition, an audio processor 130 is also coupled to the central processor 100, so that recording on the local can be enabled through a microphone 132, and so that sound stored on the local can be played through a speaker 131.
Embodiments of the present invention also provide a computer-readable program, where when the program is executed in an electronic device, the program causes a computer to execute the fund risk determination method in the electronic device according to the above embodiments.
Embodiments of the present invention further provide a storage medium storing a computer-readable program, where the computer-readable program enables a computer to execute the fund risk determination described in the above embodiments in an electronic device.
The fund risk determining scheme provided by the embodiment of the invention overcomes the limitation that the prior internet fund products mainly adopt a qualitative method, constructs a new internet fund product market risk prediction model, overcomes the defects of time consumption and high cost in artificial neural network training, constructs a fluctuation rate prediction model mixing a GARCH model and a machine learning model to predict the risk of the fund product, and forms risk prevention rating through an intelligent retrieval system so as to make auxiliary reference for users to purchase fund products.
The preferred embodiments of the present invention have been described above with reference to the accompanying drawings. The many features and advantages of the embodiments are apparent from the detailed specification, and thus, it is intended by the appended claims to cover all such features and advantages of the embodiments that fall within the true spirit and scope thereof. Further, since numerous modifications and changes will readily occur to those skilled in the art, it is not desired to limit the embodiments of the invention to the exact construction and operation illustrated and described, and accordingly, all suitable modifications and equivalents may be resorted to, falling within the scope thereof.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The principle and the implementation mode of the invention are explained by applying specific embodiments in the invention, and the description of the embodiments is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.

Claims (13)

1. A fund risk determination method, comprising:
obtaining basic metal data;
respectively carrying out model training on a pre-established exponential conditional variance model and a machine learning model by utilizing the basic metal data to determine an exponential conditional variance prediction model and a machine learning prediction model;
and generating a fund risk determination result by using the index condition variance prediction model and the machine learning prediction model determined by the basic metal data and training.
2. The method of fund risk determination according to claim 1, wherein the base metal data comprises: fund yield data, fund scale data, stock market closing price data, generalized currency supply amount data and national debt index data.
3. The fund risk determination method according to claim 1, wherein the exponential conditional variance model comprises: EGARCH model;
the machine learning model comprises: EFKOS-GELM model.
4. The method of fund risk determination according to claim 3, wherein the model training of a pre-established exponential conditional variance model and machine learning model using the basic metal data to determine an exponential conditional variance prediction model and machine learning prediction model comprises:
performing initialization training on a pre-established EFKOS-GELM model by using the attribute data to determine an initialization training model;
and performing sequential learning training on the initialization training model by using the attribute data to determine an EFKOS-GELM prediction model.
5. The method of fund risk determination according to claim 1, wherein said generating a fund risk determination using said exponential conditional variance prediction model determined from the basic metallic data and training and a machine learning prediction model comprises:
generating a fund fluctuation rate predicted value of an exponential conditional variance model by using the basic metal data and the exponential conditional variance prediction model determined by training;
generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training;
and generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential condition variance model and the fund fluctuation rate predicted value of the machine learning model.
6. The fund risk determination method according to claim 5, wherein generating the fund volatility prediction value for the machine learning model using the machine learning prediction model determined from the base metal data and training comprises:
performing principal component analysis processing on the basic metal data to generate prediction data;
and generating a fund fluctuation rate predicted value of the machine learning model by using the generated predicted data and the machine learning predicted model determined by training.
7. The fund risk determination method according to claim 5, wherein the generating a fund risk determination result based on the fund volatility predicted value of the exponential conditional variance model and the fund volatility predicted value of the machine learning model comprises:
and mixing the fund fluctuation rate predicted value of the exponential conditional variance model and the fund fluctuation rate predicted value of the machine learning model according to a preset weighted mixing proportion to generate a fund risk determination result.
8. A fund risk determination apparatus, the apparatus comprising:
the data acquisition module is used for acquiring basic metal data;
the model training module is used for respectively carrying out model training on a pre-established exponential conditional variance model and a machine learning model by utilizing the basic metal data to determine an exponential conditional variance prediction model and a machine learning prediction model;
and the risk determination module is used for generating a fund risk determination result by utilizing the basic metal data and the index condition variance prediction model and the machine learning prediction model determined by training.
9. The fund risk determination apparatus according to claim 8, wherein the exponential conditional variance model comprises: EGARCH model;
the machine learning model comprises: EFKOS-GELM model.
10. The fund risk determination device according to claim 8, wherein the model training module comprises:
the initialization unit is used for performing initialization training on a pre-established EFKOS-GELM model by utilizing the attribute data to determine an initialization training model;
and the EFKOS-GELM prediction model determining unit is used for performing sequential learning training on the initialization training model by using the attribute data to determine the EFKOS-GELM prediction model.
11. The fund risk determination device of claim 8, wherein the risk determination module comprises:
the first prediction fluctuation rate determining unit is used for generating a fund fluctuation rate prediction value of an exponential conditional variance model by using the basic metal data and the exponential conditional variance prediction model determined by training;
the second prediction fluctuation rate determining unit is used for generating a fund fluctuation rate predicted value of the machine learning model by using the basic metal data and the machine learning prediction model determined by training;
and the mixing unit is used for generating a fund risk determination result according to the fund fluctuation rate predicted value of the exponential condition variance model and the fund fluctuation rate predicted value of the machine learning model.
12. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method of any of claims 1 to 7 when executing the computer program.
13. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program for executing the method of any one of claims 1 to 7.
CN202110515484.3A 2021-05-12 2021-05-12 Fund risk determining method and device Pending CN113516559A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110515484.3A CN113516559A (en) 2021-05-12 2021-05-12 Fund risk determining method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110515484.3A CN113516559A (en) 2021-05-12 2021-05-12 Fund risk determining method and device

Publications (1)

Publication Number Publication Date
CN113516559A true CN113516559A (en) 2021-10-19

Family

ID=78064410

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110515484.3A Pending CN113516559A (en) 2021-05-12 2021-05-12 Fund risk determining method and device

Country Status (1)

Country Link
CN (1) CN113516559A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114036208A (en) * 2021-11-09 2022-02-11 建信金融科技有限责任公司 Model training and sensitivity analysis method, device, equipment and medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150088783A1 (en) * 2009-02-11 2015-03-26 Johnathan Mun 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
CN108665175A (en) * 2018-05-16 2018-10-16 阿里巴巴集团控股有限公司 A kind of processing method, device and the processing equipment of insurance business risk profile
CN108985489A (en) * 2018-06-08 2018-12-11 阿里巴巴集团控股有限公司 A kind of Risk Forecast Method, risk profile device and terminal device
WO2019061976A1 (en) * 2017-09-28 2019-04-04 平安科技(深圳)有限公司 Fund product recommendation method and apparatus, terminal device, and storage medium

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150088783A1 (en) * 2009-02-11 2015-03-26 Johnathan Mun 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
WO2019061976A1 (en) * 2017-09-28 2019-04-04 平安科技(深圳)有限公司 Fund product recommendation method and apparatus, terminal device, and storage medium
CN108665175A (en) * 2018-05-16 2018-10-16 阿里巴巴集团控股有限公司 A kind of processing method, device and the processing equipment of insurance business risk profile
CN108985489A (en) * 2018-06-08 2018-12-11 阿里巴巴集团控股有限公司 A kind of Risk Forecast Method, risk profile device and terminal device

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
库宇: "集成带遗忘机制和卡尔曼滤波的OS-GELM模型及余额宝风险预测研究", 中国优秀硕士学位论文全文数据库信息科技辑, no. 2, 15 February 2021 (2021-02-15), pages 140 - 406 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114036208A (en) * 2021-11-09 2022-02-11 建信金融科技有限责任公司 Model training and sensitivity analysis method, device, equipment and medium

Similar Documents

Publication Publication Date Title
Kripfganz et al. ardl: Estimating autoregressive distributed lag and equilibrium correction models
US7283982B2 (en) Method and structure for transform regression
CN111783873B (en) User portrait method and device based on increment naive Bayes model
US11501107B2 (en) Key-value memory network for predicting time-series metrics of target entities
Archakov et al. A multivariate realized GARCH model
CN110796261A (en) Feature extraction method and device based on reinforcement learning and computer equipment
CN114386580A (en) Decision model training method and device, decision method and device, electronic equipment and storage medium
Mesters et al. Generalized dynamic panel data models with random effects for cross-section and time
CN115375471A (en) Stock market quantification method based on adaptive feature engineering
CN113516559A (en) Fund risk determining method and device
CN113407862B (en) Sequential social contact recommendation method, system and storage medium based on door mechanism
CN111652668B (en) Block chain supply chain financial system
Amiri et al. Density estimation over spatio-temporal data streams
CN114612231A (en) Stock quantitative trading method and device, terminal device and readable storage medium
Zhan et al. Model averaging for semiparametric varying coefficient quantile regression models
Khabibullin et al. Fast Estimation of Bayesian State Space Models Using Amortized Simulation-Based Inference
CN111932018B (en) Bank business performance contribution information prediction method and device
Yümlü et al. Bayesian changepoint and time-varying parameter learning in regime switching volatility models
Gächter et al. A tale of two tails: 130 years of growth-at-risk
CN111383028B (en) Prediction model training method and device, prediction method and device
Nguyen Bayesian inference for deriving some stochastic process
Michaelides et al. A Bayesian Signals Approach for the Detection of Crises
Ashwin Essays in macroeconomics and machine learning
Kohns et al. The ARR2 prior: flexible predictive prior definition for Bayesian auto-regressions
CN117078292A (en) Feedback information analysis method and device and electronic equipment

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