WO2019019346A1 - 资产配置策略获取方法、装置、计算机设备和存储介质 - Google Patents

资产配置策略获取方法、装置、计算机设备和存储介质 Download PDF

Info

Publication number
WO2019019346A1
WO2019019346A1 PCT/CN2017/103954 CN2017103954W WO2019019346A1 WO 2019019346 A1 WO2019019346 A1 WO 2019019346A1 CN 2017103954 W CN2017103954 W CN 2017103954W WO 2019019346 A1 WO2019019346 A1 WO 2019019346A1
Authority
WO
WIPO (PCT)
Prior art keywords
management product
wealth management
historical
trend
price
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.)
Ceased
Application number
PCT/CN2017/103954
Other languages
English (en)
French (fr)
Inventor
窦宏辰
马文利
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.)
OneConnect Financial Technology Co Ltd Shanghai
Original Assignee
OneConnect Financial Technology Co Ltd Shanghai
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 OneConnect Financial Technology Co Ltd Shanghai filed Critical OneConnect Financial Technology Co Ltd Shanghai
Priority to SG11201913398QA priority Critical patent/SG11201913398QA/en
Publication of WO2019019346A1 publication Critical patent/WO2019019346A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0499—Feedforward networks
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION 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
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/06—Asset management; Financial planning or analysis
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/09—Supervised learning

Definitions

  • the present application relates to the field of computer technologies, and in particular, to an asset allocation policy acquisition method, apparatus, computer device, and storage medium.
  • the inventors of the present application have found that the prior art has the following problems: the user's financial management experience is limited, and the accuracy of the user's judgment of the rise and fall of the wealth management product is low, and Due to the large number of wealth management products and the large amount of information on wealth management products, it takes a lot of time for users to view one by one on the terminal, which wastes computer network resources.
  • an asset configuration policy acquisition method is provided.
  • An asset allocation policy acquisition method includes:
  • An asset allocation strategy acquisition device includes:
  • a level obtaining unit configured to acquire a wealth management product grade of the target wealth management product according to current attribute information of the target wealth management product and a preset wealth management product level model, wherein the wealth management product level model is based on historical attribute information of the first wealth management product and the Pre-training the model of the wealth management product corresponding to the historical attribute information;
  • a trend tendency obtaining unit configured to acquire a trend tendency of the target wealth management product according to a current indicator state of the target wealth management product and a preset wealth management product trend model, wherein the wealth management product trend model is based on the second wealth management product at a historical moment
  • the state of the historical indicator and the historical trend corresponding to the historical moment are pre-trained by the model
  • the policy obtaining unit is configured to acquire an asset allocation policy according to the wealth management product level of the target wealth management product and the trend tendency of the target wealth management product.
  • a computer apparatus comprising a memory and a processor, the memory storing computer readable instructions, the computer readable instructions being executed by the processor such that the processor performs the following steps:
  • the wealth management product grade is pre-trained by the model
  • the trend of the target wealth management product tends to be pre-trained according to the historical indicator state of the second wealth management product at the historical moment and the historical trend corresponding to the historical moment;
  • One or more non-transitory readable storage mediums storing computer readable instructions, when executed by one or more processors, cause the one or more processors to perform the following steps:
  • the wealth management product grade is pre-trained by the model
  • FIG. 1 is an implementation environment diagram of an asset configuration policy acquisition method provided in an embodiment
  • FIG. 2 is a block diagram showing the internal structure of a computer device in an embodiment
  • FIG. 3 is a flowchart of an asset allocation policy acquisition method in an embodiment
  • FIG. 5 is a flowchart of an asset configuration policy acquisition method in an embodiment
  • FIG. 6 is a structural block diagram of an asset allocation policy acquisition apparatus in an embodiment
  • FIG. 7 is a structural block diagram of an asset allocation policy acquisition apparatus in an embodiment
  • FIG. 8 is a structural block diagram of an asset configuration policy obtaining apparatus in an embodiment.
  • first may be used to describe various elements, and the elements are not limited by these terms unless specifically stated. These terms are only used to distinguish one element from another.
  • the first preset threshold may be referred to as a second preset threshold without departing from the scope of the present application, and similarly, the second preset threshold may be referred to as a first preset threshold.
  • the present application clearly indicates that the above first and second are indicative of the order, for example, herein, the first, nth, and tth include the limitation of the order.
  • FIG. 1 is an implementation environment diagram of an asset configuration policy acquisition method provided in an embodiment. As shown in FIG. 1 , the implementation environment includes a market feature acquisition device 110, a computer device 120, and a transaction device 130.
  • the market feature device 110 obtains the current wealth management product market of the target wealth management product, and inputs the result to the computer device 120.
  • the computer device 120 according to the current wealth management product market, the current attribute information of the target wealth management product and the wealth management product level model.
  • Obtain the wealth management product grade of the target wealth management product obtain the trend tendency of the target wealth management product according to the current indicator status of the target wealth management product and the wealth management product trend model, and then obtain the asset allocation strategy according to the wealth management product grade of the target wealth management product and the trend tendency, when obtained
  • the computer device 120 can output the asset configuration policy to the client for viewing by the user. You can also send transactions directly according to the asset allocation policy.
  • the transaction device 130 is requested to conduct a transaction.
  • the user may be prompted to purchase the wealth management product, and the transaction device may be sent a transaction request to purchase the wealth management product, and the transaction volume may be determined according to the user preset. It can also be judged according to the specific grade of the wealth management product and the intensity of the trend tendency.
  • the above-mentioned computer device 120 may be an independent physical server or terminal, or may be a server cluster composed of multiple physical servers, and may be a cloud server that provides basic cloud computing services such as a cloud server, a cloud database, a cloud storage, and a CDN.
  • Wealth management products can be precious metals such as silver and gold, as well as oil, stocks and futures.
  • FIG. 2 is an internal structural diagram of a computer device in one embodiment
  • the computer device is connected to a processor, a non-volatile storage medium, an internal memory, and a network interface through a system connection bus.
  • the non-volatile storage medium of the computer device can store an operating system and computer readable instructions that, when executed, can cause the processor to perform an asset configuration policy acquisition method.
  • the processor of the computer device is used to provide computing and control capabilities to support the operation of the entire computer device.
  • the internal memory can store computer readable instructions that, when executed by the processor, cause the processor to perform an asset configuration policy acquisition method.
  • the network interface of the computer device is used for network communication, such as sending an asset configuration policy. It will be understood by those skilled in the art that the structure shown in FIG.
  • FIG. 2 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation of the computer device to which the solution of the present application is applied.
  • the specific computer device may It includes more or fewer components than those shown in the figures, or some components are combined, or have different component arrangements.
  • an asset configuration policy acquisition method is provided.
  • the asset configuration policy acquisition method may be applied to the foregoing computer device 120, and specifically includes the following steps:
  • Step S302 Acquire a wealth management product grade of the target wealth management product according to the current attribute information of the target wealth management product and the preset wealth management product level model.
  • the property information of wealth management products is the essential characteristics or characteristics of wealth management products, such as the price-earnings ratio, market rate, current ratio and information ratio of wealth management products.
  • the current attribute information refers to the attribute information of the target wealth management product at the current time
  • the historical attribute information refers to the genus of the first wealth management product at the historical moment before the current time.
  • the number of attribute information of the current time and historical time wealth management products can be set according to actual needs.
  • the attribute information of the current time and historical time wealth management products may be two price-to-earning ratios and a market current rate. It can also be four price-to-earnings ratios, market rates, current ratios, and information ratios.
  • the wealth management product level corresponding to the historical attribute information is the level of the first wealth management product under the historical attribute information.
  • the level of wealth management products can be quantified by pre-set levels such as excellent, medium and poor, or can be quantified by specific scores, as long as the wealth management products can be classified.
  • the preset wealth management product level model is obtained by pre-training according to the historical attribute information of the first wealth management product and the wealth management product level corresponding to the historical attribute information. After training the model, the current attribute information of the target wealth management product is input to the model. In the trained wealth management product level model, the wealth management product grade corresponding to the target wealth management product can be obtained.
  • Step S304 Obtain a trend tendency of the target wealth management product according to the current indicator status of the target wealth management product and the preset wealth management product trend model.
  • the indicator status refers to one or a combination of the status of each indicator of the wealth management product and the relationship between the indicators.
  • it may be a size relationship between moving averages of wealth management products, whether DEA (Difference Exponential Average) is positive or negative, DIF (Difference, disparity) and DEA (Difference Exponential Average, smoothing)
  • the relationship between the indicators can be divided into multiple, for example, the index 1 is greater than the index 2, the index 1 is equal to the index 2, the index 1 is smaller than the index 2, the index 1 is equal to the index 2, and the value of the index 1 is decreased.
  • the 5-day moving average is greater than the 10-day moving average
  • the DIF is higher than DEA from low to lower, and the distance between the stock pressure line, the support line, and the stock average is narrowed.
  • the current indicator status refers to the indicator status of the target wealth management product at the current moment.
  • the historical indicator status refers to the status of the indicator of the second wealth management product at the historical moment before the current time. It can be understood that the time can refer to a time point or a period of time.
  • the indicator status of the current time and the number of indicator status of the historical time can be set according to the actual situation.
  • the historical trend corresponding to the historical moment refers to the trend of the second wealth management product relative to the historical moment in the next period of history, for example, a historical moment is 2016. On December 15th, the corresponding historical trend can be the ups and downs after 3 days, that is, on December 18, 2016, compared with December 15, the price of wealth management products is rising, falling or oscillating.
  • Trends tend to refer to the ups and downs of wealth management products, which can be rise, fall and shock, or the probability of a rising or falling trend, for example, the probability of rising is 80%.
  • Concussion refers to the unstable price of wealth management products, which is high when rising.
  • the preset wealth management product trend model is based on the historical indicator status of the second wealth management product at the historical moment and the historical trend corresponding to the historical moment. The model is pre-trained after the model is trained, and the current indicator status of the target wealth management product is input into the training. In the good wealth management product trend model, the trend tendency of the target wealth management products can be obtained.
  • Step S306 obtaining an asset allocation strategy according to the wealth management product grade of the target wealth management product and the trend tendency of the target wealth management product.
  • An asset allocation strategy is a proposal to buy or sell wealth management products and can include one or more wealth management products. For example, for a wealth management product held by a user, if it is judged that the level of the wealth management product is low, and the trend of the wealth management product tends to fall, it may be recommended to sell the wealth management product. For the wealth management products that the user does not hold, if the user's stock selection conditions are met, and the level is high and the trend is rising, the user may be sent an asset allocation strategy for buying the wealth management product. Of course, for the wealth management products that the user already holds, and the level is high and the trend is rising, the user may be advised to continue to buy.
  • the asset allocation strategy may also include a wealth management product transaction volume.
  • the trading volume of wealth management products can be determined according to the trend tendency of the target wealth management products and/or the level of wealth management products. For example, set the correspondence between the trend tendency and the transaction volume, and the correspondence between the wealth management product level and the wealth management product transaction volume. The higher the trend is, the higher the probability of a rise, the greater the volume of transactions bought, and the higher the level of wealth management products, the greater the volume of transactions bought.
  • the corresponding wealth management product transaction volume may be obtained according to the correspondence between the trend tendency and the transaction volume and/or the correspondence between the wealth management product level and the wealth management product transaction volume.
  • the quantity of the first wealth management product, the second wealth management product and the target wealth management product may be one or more, and the first wealth management product, the second wealth management product and the target wealth management product may be completely different or identical. , or partially the same.
  • the method for obtaining the target wealth management product is not limited in the embodiment of the present application.
  • the method may be acquired according to the type of risk that the user can bear, or may be randomly obtained. It is also possible to select an industry wealth management product as a target wealth management product according to the user's choice of an industry.
  • steps S302 and S304 may be performed simultaneously or sequentially.
  • the execution order of S302 and S304 may be set according to actual needs, and the asset policy configuration apparatus receives and executes steps S302 and S304 according to the execution order.
  • the trend tendency of the target wealth management product may be first obtained, and when the trend tends to rise, the level of the target wealth management product may be acquired.
  • the level of the target wealth management product is first obtained, and when the level is higher than the preset level, it is the trend of the target wealth management product.
  • the step S304 is to obtain a trend tendency of the target wealth management product according to the current indicator status of the target wealth management product and the preset wealth management product trend model, including: when the wealth management product level of the target wealth management product is greater than or equal to the preset level.
  • the trend tendency of the target wealth management product is obtained.
  • Step S306 The step of acquiring an asset allocation strategy according to the level of the wealth management product and the trend tendency includes: when the wealth management product grade of the target wealth management product is greater than or equal to the preset level, and the trend of the target wealth management product tends to rise, obtaining the target financial product for purchase Asset allocation strategy.
  • the wealth management product grade Take the wealth management product grade as the excellent, medium and bad grades. If the preset grade is medium, when the wealth management product grade obtained according to the wealth management product grade model is medium or excellent, the current indicator status of the target wealth management product is input to In the preset wealth management product trend model, the trend tendency of obtaining the target wealth management product is obtained, and if the trend tends to rise, the asset allocation strategy for buying the target wealth management product is obtained. If the level of the wealth management product of the target wealth management product is poor, the trend of the trend of the wealth management product is not predicted.
  • the above-mentioned asset allocation strategy acquisition method is based on the historical property information of the first wealth management product and the wealth management product level corresponding to the historical property information, and the wealth management product trend model is based on the history of the second wealth management product.
  • the status of the indicator and the historical trend corresponding to the historical moment are pre-trained by the model. Since the attribute information of the wealth management product and the status of the indicator often contain the rank of the wealth management product and the law of the subsequent rise of the wealth management product, the historical wealth management product attribute information is utilized. And the indicator status information for model training can accurately find out the law of these data.
  • the preset wealth management product level model predicts the wealth management product level of the wealth management product, predicts the trend tendency of the wealth management product according to the current indicator state of the target wealth management product and the preset wealth management product trend model, and then according to the wealth management product grade of the target wealth management product and the wealth management product.
  • the trend tends to obtain the corresponding asset allocation strategy, which saves the time for users to view a large amount of information, and the acquired asset allocation strategy has a high accuracy rate.
  • the asset policy configuration method may further include the following steps:
  • the historical price time series is a sequence of historical time and the price of the preset time after the historical time are sorted in order of time development.
  • the preset time can be set according to actual requirements.
  • the preset time after the historical moment includes the historical moment.
  • the time interval between each historical price is the same. For example, it can be a daily closing price of 7 consecutive days.
  • the historical time is May 1, 2016, and the preset time is 7 days.
  • the historical price time series is a sequence of prices for each day from May 1 to May 8, for a total of 8 prices.
  • Price historical moment is P 0, i.e. initial price P 0, the final price of P n
  • the historical price series can be expressed as ⁇ P 0, P 1, P 2, ...... P n ⁇ .
  • S404 Calculate a trend judgment parameter according to a historical price time series, where the trend judgment parameter includes a price standard score, or the trend judgment parameter includes a price standard score and a change utility coefficient.
  • the price standard score is the difference between the n+1th price in the historical price time series minus the historical price time series average value divided by the historical price time series standard deviation
  • the change utility coefficient is the tth price in the historical price series
  • the absolute value of the difference from the first price is then divided by the sum of the absolute values of the adjacent price differences between the first price and the tth price, t is equal to n or n+1, and n is a positive integer.
  • the n+1th price refers to the final price in the historical price series
  • the first price refers to the initial price in the historical price series, which is expressed by a formula
  • the formula of the price standard score A can be as shown in the formula (1).
  • the formula for changing the utility coefficient B can be as shown in formula (2).
  • P is the historical price series
  • P n is the n+1th price, that is, the final price in the preset time
  • P 0 is the first price, that is, the price at the historical time is the starting price
  • P i+1 For the i+2th price in the historical price series, P i is the i+1th price in the historical price series, where i is greater than or equal to 0 and less than or equal to n-1.
  • E(P) is the mean of the historical price series
  • ⁇ (P) is the standard deviation of the historical price series.
  • Equation 2 can be expressed as formula (3), which is known from Equation 3.
  • B is less than or equal to 1 and greater than or equal to 0. If it is equal to 1, it indicates that the wealth management product has been rising or falling all the time. When there are other values, there is fluctuation, so the fluctuation of the wealth management products can be judged according to the size of B.
  • the historical trend corresponding to the historical moment refers to the trend of the second wealth management product relative to the historical moment in the next period of history.
  • the trend judgment parameter includes the price standard score
  • the historical trend corresponding to the historical moment is increased when the price standard score is greater than the preset threshold, and the historical trend corresponding to the historical moment is decreased when the threshold is less than the preset threshold.
  • the trend judgment parameter includes the price standard score and the change utility coefficient
  • the historical trend can be determined by combining the price standard score and the change utility coefficient.
  • the step of obtaining the historical trend corresponding to the historical moment includes: when the price standard score is greater than the first preset threshold, and the change utility coefficient is greater than the second preset threshold, the historical trend corresponding to the historical moment is increased; when the absolute value of the price standard is When the value is less than the first preset threshold, and the change utility coefficient is less than the second preset threshold, the historical trend corresponding to the historical moment is obtained as an oscillation; when the price standard score is less than the first preset threshold, and the change utility coefficient is greater than the second pre- When the threshold is set, the historical trend corresponding to the historical moment is decreased; wherein the first preset threshold and the second preset threshold are greater than 0 and less than or equal to 1.
  • the first preset threshold is 0.1 and the second preset threshold is 0.4
  • the price standard score of a second wealth management product is 0.5 and the change utility coefficient is 0.6
  • the historical trend is increased.
  • the acquired trend sample model predicts good results.
  • the historical indicators of the second wealth management product in the historical moment and the historical trend corresponding to the historical moment constitute the training data, and the model training is performed according to the training data, and the trend model of the wealth management product is obtained.
  • the number of the second wealth management products may be one or more, and the number of the training data may be specifically obtained according to actual conditions, for example, several hundred or tens of thousands.
  • Each training data includes the historical indicator status of a second wealth management product at a historical moment and the corresponding historical trend.
  • the historical indicator status can be one or more. For example, according to the historical time at 12:55 on June 16, 2016, the historical indicator status may include a 5-day moving average greater than a 10-day moving average, a DEA value greater than 0, and a DEA value from a small value. Going up above the DIF value, etc., the corresponding historical trend is rising.
  • the machine learning is performed according to the training data, and the model parameters obtained by the machine learning training are obtained, and the wealth management product trend model is obtained.
  • the model of machine learning can be Support Vector Machine (SVM) classifier model, Artificial Neural Network (ANN) classifier model, Logistic Regression (LR) classifier model and hidden Markov.
  • SVM Support Vector Machine
  • ANN Artificial Neural Network
  • LR Logistic Regression
  • Various models for classification such as the model (Hidden Markov Model, HMM).
  • the kernel function used can be set according to actual requirements.
  • the support vector machine can be used for supervised machine learning, and the kernel function can adopt a polynomial function.
  • the asset policy configuration method may further include the following steps:
  • the number of the first wealth management products may be one or more.
  • the number of the first training samples in the first training sample set may be specifically obtained according to actual conditions, for example, several hundred or tens of thousands.
  • Each first training sample includes a plurality of historical attribute information and a corresponding wealth management product level.
  • the historical attribute information may include a price-to-earnings ratio of 5%, a market rate of 10%, a turnover rate of 20%, etc.
  • the corresponding wealth management product grade is a blue-chip stock.
  • the level of wealth management products can be marked by humans, or by obtaining data of wealth management products, and then obtained according to relevant quantitative formulas.
  • the level of the first wealth management product can be judged by the Sotino ratio, which is the unit income that can be obtained for each unit of downward fluctuation. The larger the value, the risk of the same decline. In the case of the situation, you can get more excess returns or returns on the benchmark.
  • Sotino ratio is the unit income that can be obtained for each unit of downward fluctuation. The larger the value, the risk of the same decline. In the case of the situation, you can get more excess returns or returns on the benchmark.
  • it may be set as a non-preferred stock when the Sotino ratio is greater than the third preset threshold, and is a good stock, and less than a certain fourth preset threshold.
  • the third preset threshold and the fourth preset threshold may be set according to specific requirements.
  • S504 Perform model training according to the first training sample set to obtain a wealth management product level model.
  • the model of machine learning can be Support Vector Machine (SVM) classifier model, Artificial Neural Network (ANN) classifier model, Logistic Regression (LR) classifier model and hidden Markov.
  • SVM Support Vector Machine
  • ANN Artificial Neural Network
  • LR Logistic Regression
  • Various models for classification such as the model (Hidden Markov Model, HMM).
  • the kernel function used can be set according to actual requirements.
  • the support vector machine can be used for supervised machine learning, and the kernel function can adopt a polynomial function.
  • an asset configuration policy obtaining device is provided.
  • the asset configuration policy obtaining device may be integrated into the computer device 120, and may include a level obtaining unit 602 and a trend tendency obtaining unit. 604 and a policy acquisition unit 606.
  • a level obtaining unit 602 configured to use current attribute information of the target wealth management product and preset financial management
  • the product level model obtains the wealth management product grade of the target wealth management product.
  • the preset wealth management product level model is obtained by pre-training the model according to the historical attribute information of the first wealth management product and the wealth management product level corresponding to the historical attribute information.
  • the trend tendency obtaining unit 604 is configured to obtain a trend tendency of the target wealth management product according to the current indicator state of the target wealth management product and the preset wealth management product trend model.
  • the preset wealth management product trend model is obtained by pre-training the model according to the historical indicator status of the second wealth management product at the historical moment and the historical trend corresponding to the historical moment.
  • the policy obtaining unit 606 is configured to obtain an asset allocation policy according to the wealth management product level of the target wealth management product and the trend tendency of the target wealth management product.
  • the trend tendency obtaining unit 604 is configured to: when the wealth management product level of the target wealth management product is greater than or equal to the preset level, obtain the target wealth management product according to the current indicator state of the target wealth management product and the preset wealth management product trend model. The trend is trending.
  • the policy obtaining unit 606 is configured to: when the wealth management product level of the target wealth management product is greater than or equal to the preset level, and the trend of the target wealth management product is inclined to increase, obtain an asset allocation strategy for buying the target wealth management product.
  • the wealth management product level model is obtained by pre-training the model according to the historical attribute information of the first wealth management product and the wealth management product level corresponding to the historical attribute information, and the wealth management product trend model is based on the history of the second wealth management product.
  • the status of the indicator and the historical trend corresponding to the historical moment are pre-trained by the model.
  • the attribute information and indicator status of the wealth management product often contain the level of the wealth management product and the follow-up trend of the wealth management product, so the historical wealth management product attribute information is used.
  • Model status information for model training can accurately find out the laws of these data.
  • the wealth management product level of the wealth management product can be predicted according to the current attribute information of the target wealth management product and the preset wealth management product level model, according to the current indicator status of the target wealth management product and the preset wealth management product trend.
  • the model predicts the trend tendency of wealth management products, and then obtains the corresponding asset allocation strategy according to the wealth management product grade of the target wealth management product and the trend tendency of the wealth management product, which saves the time for the user to view a large amount of information, and has high accuracy.
  • the asset policy configuration apparatus may further include a price time series acquisition unit 702, a trend determination parameter calculation unit 704, a history trend acquisition unit 706, and a second Model training unit 708:
  • the price time series obtaining unit 702 is configured to obtain a historical price time series of the preset time of the second wealth management product after the historical time, and the historical price time series includes n+1 prices.
  • the trend judgment parameter calculation unit 704 is configured to acquire a trend judgment parameter of the historical price time series, the trend judgment parameter includes a price standard score, or the trend judgment parameter includes a price standard score and a change utility coefficient.
  • the historical trend obtaining unit 706 is configured to obtain a historical trend according to the trend determining parameter and the preset determining rule.
  • the second model training unit 708 is configured to compose the training data of the historical indicator state of the second wealth management product at the historical moment and the historical trend corresponding to the historical moment, and perform model training according to the training data to obtain a wealth management product trend model.
  • the asset policy configuration apparatus may further include a first sample set acquisition unit 802 and a first model training unit 804:
  • the first sample set obtaining unit 802 is configured to acquire a first training sample set that is composed of a plurality of first training samples, where the first training sample includes multiple historical attribute information of the first wealth management product at a historical moment and a historical attribute corresponding to Financial product grade;
  • the first model training unit 804 is configured to perform model training according to the first training sample set to obtain a wealth management product level model.
  • the network interface may be an Ethernet card or a wireless network card.
  • the above modules may be embedded in the hardware in the processor or in the memory in the server, or may be stored in the memory in the server, so that the processor calls the corresponding operations of the above modules.
  • the processor can be a central processing unit (CPU), a microprocessor, a microcontroller, or the like.
  • the asset configuration policy obtaining apparatus may be implemented in the form of a computer readable instruction executable on a computer device as shown in FIG. 2, a nonvolatile of the computer device
  • the storage medium may store various program modules constituting the asset configuration policy acquisition means, such as the level acquisition unit 602, the trend tendency acquisition unit 604, and the policy acquisition in FIG. Unit 606.
  • Computer program readable instructions are included in each program module for causing a computer device to perform the steps in the asset configuration policy acquisition method of various embodiments of the present application described in this specification.
  • the computer device may acquire the wealth management product level of the target wealth management product according to the current attribute information of the target wealth management product and the preset wealth management product level model, as shown in FIG.
  • the trend tendency acquisition unit 604 by the trend tendency acquisition unit 604 according to the target.
  • the current indicator status of the wealth management product and the preset wealth management product trend model obtain the trend tendency of the target wealth management product
  • the strategy acquisition unit 606 obtains the asset allocation strategy according to the wealth management product level of the target wealth management product and the trend tendency of the target wealth management product.
  • a computer device is proposed.
  • the internal structure of the computer device may correspond to the structure shown in FIG. 2, that is, the computer device may be a server or a terminal, and the computer device includes a memory, a processor, and Computer readable instructions stored on the memory and operable on the processor, the processor executing the computer readable instructions to: obtain the target wealth management product according to the current attribute information of the target wealth management product and the preset wealth management product level model
  • the wealth management product level and the wealth management product level model are pre-trained according to the historical attribute information of the first wealth management product and the wealth management product level corresponding to the historical attribute information; the target is obtained according to the current indicator state of the target wealth management product and the preset wealth management product trend model.
  • the trend of wealth management products tends to be based on the historical indicators of the second wealth management products in the historical moments and the historical trends corresponding to historical moments.
  • the model is pre-trained according to the wealth management products of the target wealth management products and the target wealth management products. Trends tend to acquire asset allocation strategies.
  • the processor when executing the computer readable instructions, further performs the steps of: obtaining a historical price time series of the second financial product at a preset time after the historical time, the historical price time series including n+1 prices;
  • the price time series calculation obtains the trend judgment parameter, the trend judgment parameter includes the price standard score, or the trend judgment parameter includes the price standard score and the change utility coefficient;
  • the historical trend corresponding to the historical moment is obtained according to the trend judgment parameter;
  • the second wealth management product is in the historical moment
  • the historical indicator state and the historical trend corresponding to the historical moment constitute the training data, and the model training is performed according to the training data to obtain the wealth management product trend model;
  • the price standard score includes the n+1th price minus the historical price in the historical price time series
  • the difference between the time series mean values is divided by the standard deviation of the historical price time series
  • the change utility coefficient is the difference between the tth price and the first price in the historical price series.
  • the absolute value is then divided by the sum of the absolute values of the
  • the historical trend corresponding to the historical moment is obtained according to the trend determining parameter, including: when the price standard score is greater than the first preset threshold, and the changing utility coefficient is greater than the second preset threshold, obtaining a historical trend corresponding to the historical moment When the absolute value of the price standard score is less than the first preset threshold, and the utility coefficient is less than the second preset threshold, the historical trend corresponding to the historical moment is obtained as an oscillation; when the price standard score is less than the first preset threshold, When the utility coefficient is greater than the second preset threshold, the historical trend corresponding to the historical time is decreased; wherein the first preset threshold and the second preset threshold are greater than 0 and less than or equal to 1.
  • the trend tendency of the target wealth management product is obtained according to the current indicator status of the target wealth management product and the preset wealth management product trend model, including: when the wealth management product level of the target wealth management product is greater than or equal to the preset level, according to the target
  • the current indicator status of the wealth management product and the preset wealth management product trend model obtain the trend tendency of the target wealth management product
  • the steps of obtaining the asset allocation strategy according to the wealth management product level and the trend tendency include: when the wealth management product level of the target wealth management product is greater than or equal to the preset Level, and the trend of the target wealth management products tends to increase, and obtain the asset allocation strategy of buying the target wealth management products.
  • the asset allocation strategy is obtained according to the level of the wealth management product of the target wealth management product and the trend of the target wealth management product, and further includes: determining the target wealth management product according to the trend tendency of the target wealth management product and/or the wealth management product grade of the target wealth management product. Corresponding transaction volume.
  • a storage medium storing computer readable instructions that, when executed by one or more processors, cause one or more processors to perform the steps of:
  • the current attribute information and the preset wealth management product level model obtain the wealth management product grade of the target wealth management product, and the wealth management product level model is pre-trained according to the historical attribute information of the first wealth management product and the wealth management product level corresponding to the historical attribute information;
  • the current indicator status of the target wealth management product and the preset wealth management product trend model obtain the trend tendency of the target wealth management product.
  • the wealth management product trend model performs model pre-training according to the historical indicator status of the second wealth management product at the historical moment and the historical trend corresponding to the historical moment. Get; according to the level of wealth management products of the target wealth management products and Trends in targeted wealth management products tend to capture asset allocation strategies.
  • the processor when executing the computer readable instructions, further performs the steps of: obtaining a historical price time series of the second financial product at a preset time after the historical time, the historical price time series including n+1 prices;
  • the price time series calculation obtains the trend judgment parameter, the trend judgment parameter includes the price standard score, or the trend judgment parameter includes the price standard score and the change utility coefficient;
  • the historical trend corresponding to the historical moment is obtained according to the trend judgment parameter;
  • the second wealth management product is in the historical moment
  • the historical indicator state and the historical trend corresponding to the historical moment constitute the training data, and the model training is performed according to the training data to obtain the wealth management product trend model;
  • the price standard score includes the n+1th price minus the historical price in the historical price time series
  • the difference between the time series averages is divided by the standard deviation of the historical price time series
  • the change utility coefficient is the absolute value of the difference between the tth price and the first price in the historical price series divided by the first price to the tth
  • the historical trend corresponding to the historical moment is obtained according to the trend determining parameter, including: when the price standard score is greater than the first preset threshold, and the changing utility coefficient is greater than the second preset threshold, obtaining a historical trend corresponding to the historical moment When the absolute value of the price standard score is less than the first preset threshold, and the utility coefficient is less than the second preset threshold, the historical trend corresponding to the historical moment is obtained as an oscillation; when the price standard score is less than the first preset threshold, When the utility coefficient is greater than the second preset threshold, the historical trend corresponding to the historical time is decreased; wherein the first preset threshold and the second preset threshold are greater than 0 and less than or equal to 1.
  • the trend tendency of the target wealth management product is obtained according to the current indicator status of the target wealth management product and the preset wealth management product trend model, including: when the wealth management product level of the target wealth management product is greater than or equal to the preset level, according to the target
  • the current indicator status of the wealth management product and the preset wealth management product trend model obtain the trend tendency of the target wealth management product
  • the steps of obtaining the asset allocation strategy according to the wealth management product level and the trend tendency include: when the wealth management product level of the target wealth management product is greater than or equal to the preset Level, and the trend of the target wealth management products tends to increase, and obtain the asset allocation strategy of buying the target wealth management products.
  • the asset allocation strategy is obtained according to the level of the wealth management product of the target wealth management product and the trend of the target wealth management product, and further includes: according to the trend tendency and/or target of the target wealth management product.
  • the wealth management product level of the wealth management product determines the transaction volume corresponding to the target wealth management product.
  • the storage medium may be a non-volatile storage medium such as a magnetic disk, an optical disk, or a read-only memory (ROM).

Landscapes

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

Abstract

一种资产配置策略获取方法，包括：根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取所述目标理财产品的理财产品等级，所述理财产品等级模型根据第一理财产品的历史属性信息以及所述历史属性信息对应的理财产品等级进行模型预训练得到；根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，所述理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及所述历史时刻对应的历史趋势进行模型预训练得到；根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略。

Description

资产配置策略获取方法、装置、计算机设备和存储介质
本申请要求于2017年7月25日提交中国专利局、申请号为2017106144289、发明名称为“资产配置策略获取方法、装置、计算机设备和存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及计算机技术领域，特别是涉及资产配置策略获取方法、装置、计算机设备和存储介质。
背景技术
在理财投资领域中，为了使理财资本能够获得好的收益率，需要对理财产品如股票的的涨跌进行判断，以根据判断进行理财产品买卖。目前，对理财产品的涨跌的判断主要是用户依靠查看理财产品的信息后依靠以往理财经验来对未来涨跌进行判断。
然而，在对现有技术的研究与实践中，本申请的发明人发现，现有技术存在以下问题：用户的理财经验是有限的，导致用户判断理财产品的涨跌的准确率低，而且，由于理财产品数量多且理财产品的信息量大，用户在终端上逐个查看需要花费大量的时间，浪费了计算机网络资源。
发明内容
根据本申请的各种实施例，提供一种资产配置策略获取方法、装置、计算机设备和存储介质。
一种资产配置策略获取方法，包括：
根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取所述目标理财产品的理财产品等级，所述理财产品等级模型根据第一理财产品的历 史属性信息以及所述历史属性信息对应的理财产品等级进行模型预训练得到；
根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，所述理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及所述历史时刻对应的历史趋势进行模型预训练得到；
根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略。
一种资产配置策略获取装置，包括：
等级获取单元，用于根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取所述目标理财产品的理财产品等级，所述理财产品等级模型根据第一理财产品的历史属性信息以及所述历史属性信息对应的理财产品等级进行模型预训练得到；
趋势倾向获取单元，用于根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，所述理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及所述历史时刻对应的历史趋势进行模型预训练得到；
策略获取单元，用于根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略。
一种计算机设备，包括存储器和处理器，所述存储器中存储有计算机可读指令，所述计算机可读指令被所述处理器执行时，使得所述处理器执行以下步骤：
根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取所述目标理财产品的理财产品等级，所述理财产品等级模型根据第一理财产品的历史属性信息以及所述历史属性信息对应的理财产品等级进行模型预训练得到；
根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取 所述目标理财产品的趋势倾向，所述理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及所述历史时刻对应的历史趋势进行模型预训练得到；
根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略。
一个或多个存储有计算机可读指令的非易失性可读存储介质，所述计算机可读指令被一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：
根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取所述目标理财产品的理财产品等级，所述理财产品等级模型根据第一理财产品的历史属性信息以及所述历史属性信息对应的理财产品等级进行模型预训练得到；
根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，所述理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及所述历史时刻对应的历史趋势进行模型预训练得到；
根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略。
本申请的一个或多个实施例的细节在下面的附图和描述中提出。本申请的其它特征、目的和优点将从说明书、附图以及权利要求书变得明显。
附图说明
为了更清楚地说明本申请实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为一个实施例中提供的资产配置策略获取方法的实施环境图；
图2为一个实施例中计算机设备的内部结构框图；
图3为一个实施例中资产配置策略获取方法的流程图；
图4为一个实施例中资产配置策略获取方法的流程图；
图5为一个实施例中资产配置策略获取方法的流程图；
图6为一个实施例中资产配置策略获取装置的结构框图；
图7为一个实施例中资产配置策略获取装置的结构框图；
图8为一个实施例中资产配置策略获取装置的结构框图。
具体实施方式
为了使本申请的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本申请进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。
可以理解，本申请所使用的术语“第一”、“第二”等可在本文中用于描述各种元件，除非特别说明，否则这些元件不受这些术语限制。这些术语仅用于将第一个元件与另一个元件区分。举例来说，在不脱离本申请的范围的情况下，可以将第一预设阈值称为第二预设阈值，且类似地，可将第二预设阈值称为第一预设阈值。但是，若本申请明确的表示上述第一、第二是表示次序的除外，例如，本文中，第一个、第n个、第t个包括了对顺序的限制。
图1为一个实施例中提供的资产配置策略获取方法的实施环境图，如图1所示，在该实施环境中，包括行情特征获取装置110、计算机设备120以及交易装置130。
如图1所示，行情特征装置110获取目标理财产品当前的理财产品行情后，输入到计算机设备120中，计算机设备120根据输入的理财产品行情中目标理财产品的当前属性信息以及理财产品等级模型获取目标理财产品的理财产品等级、根据目标理财产品的当前指标状态以及理财产品趋势模型获取目标理财产品的趋势倾向，然后根据目标理财产品的理财产品等级以及趋势倾向获取资产配置策略，当获取到资产配置策略后，计算机设备120可以将该资产配置策略输出到客户端中，以供用户查看。也可以直接根据该资产配置策略发送交易请 求到交易装置130中，进行交易。
例如，当获取到买入A01理财产品的资产配置策略后，可以向用户输出买入该理财产品的建议，可以向交易装置发送买入该理财产品的交易请求，交易量可以根据用户预先设置确定，也可以根据理财产品具体的等级和趋势倾向的强度判断。
上述的计算机设备120可以是独立的物理服务器或终端，也可以是多个物理服务器构成的服务器集群，可以是提供云服务器、云数据库、云存储和CDN等基础云计算服务的云服务器。理财产品可以是白银、黄金等贵金属，还可以是石油、股票、期货等。
如图2所示，为一个实施例中计算机设备的内部结构图，该计算机设备通过系统连接总线连接处理器、非易失性存储介质、内存储器和网络接口。其中，该计算机设备的非易失性存储介质可存储操作系统和计算机可读指令，该计算机可读指令被执行时，可使得处理器执行一种资产配置策略获取方法。该计算机设备的处理器用于提供计算和控制能力，支撑整个计算机设备的运行。该内存储器中可储存有计算机可读指令，该计算机可读指令被处理器执行时，可使得处理器执行一种资产配置策略获取方法。计算机设备的网络接口用于进行网络通信，如发送资产配置策略等。本领域技术人员可以理解，图2中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备的限定，具体的计算机设备可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
如图3所示，在一个实施例中，提出了一种资产配置策略获取方法，该资产配置策略获取方法可以应用于上述的计算机设备120中，具体可以包括以下步骤：
步骤S302，根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取目标理财产品的理财产品等级。
理财产品的属性信息为理财产品的本质特征或特性，例如理财产品的市盈率、市现率、流动比率以及信息比率等。当前属性信息指在当前时刻目标理财产品的属性信息，历史属性信息指当前时刻之前的历史时刻第一理财产品的属 性信息。当前时刻以及历史时刻理财产品的属性信息的数量可以根据实际需要进行设置。例如，当前时刻以及历史时刻理财产品的属性信息可以为市盈率以及市现率两个。也可以为市盈率、市现率、流动比率以及信息比率四个。当然，也还可以加上理财产品的其他属性信息，以更准确的进行预测。可以理解，时刻可以指一个时间点，也可以是一段时间。
历史属性信息对应的理财产品等级为在该历史属性信息之下第一理财产品的等级。理财产品的等级可以用优、中以及差等预设的等级进行量化、也可以用具体的分数进行量化，只要能对理财产品进行分类即可。预设的理财产品等级模型是事先根据第一理财产品的历史属性信息以及历史属性信息对应的理财产品等级进行模型预训练得到的，当训练好模型后，将目标理财产品的当前属性信息输入到训练好的理财产品等级模型中，可以得到目标理财产品对应的理财产品等级。
步骤S304，根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向。
指标状态指理财产品的各个指标的状态以及指标之间的关系的其中之一或者组合。例如可以为理财产品的移动平均值之间的大小关系，可以为DEA(Difference Exponential Average、平滑移动平均)是正值还是负值，DIF(Difference、离差值)与DEA(Difference Exponential Average、平滑移动平均)的大小关系，股价与移动平均值的关系，股价压力线、支撑线以及股价平均线这三个指标的关系等。各个指标之间的关系可以分为多个，例如指标1大于指标2、指标1与指标2相等、指标1小于指标2、指标1与指标2相等且指标1的值是降低的等。例如，5日移动平均值大于10日移动平均值，DIF从低往下超过DEA，股价压力线、支撑线以及股价平均线这三者之间的距离变窄等。
当前指标状态指在当前时刻目标理财产品的指标状态。历史指标状态指当前时刻之前的历史时刻第二理财产品的指标状态，可以理解，时刻可以指一个时间点，也可以指一段时间。当前时刻的指标状态以及历史时刻的指标状态的数量可以根据实际进行设置。历史时刻对应的历史趋势指在历史时刻接下来的一段时间第二理财产品相对于历史时刻的涨跌趋势，例如，某一历史时刻是2016 年12月15日，则其对应的历史趋势可以为3天后的涨跌情况，即2016年12月18日相对于12月15日理财产品价格是上涨、下跌还是震荡等。
趋势倾向指理财产品的涨跌趋势，可以为上涨、下跌以及震荡，也可以为涨跌趋势的概率，例如，上涨概率为80％等。震荡指理财产品价格不稳定，时升时高。预设的理财产品趋势模型是根据第二理财产品在历史时刻的历史指标状态以及历史时刻对应的历史趋势进行模型预训练得到，当训练好模型后，将目标理财产品的当前指标状态输入到训练好的理财产品趋势模型中，可以得到目标理财产品的趋势倾向。
步骤S306，根据目标理财产品的理财产品等级以及目标理财产品的趋势倾向获取资产配置策略。
资产配置策略指对理财产品进行买卖的建议，可以包括一个或多个理财产品的策略。例如，对于用户持有的理财产品，如判断是理财产品等级低，理财产品的趋势倾向为下跌时，则可以建议卖出该理财产品。对于用户未持有的理财产品，如果符合用户的选股条件，且等级高、趋势为上涨，则可以向用户发送买入该理财产品的资产配置策略。当然，对于用户已持有的理财产品，且等级高、趋势为上涨，也可以建议用户继续买入。
在一些实施例中，资产配置策略还可以包括理财产品交易量。可以根据目标理财产品的趋势倾向和/或理财产品等级确定理财产品交易量。例如，设置趋势倾向与交易量的对应关系，理财产品等级与理财产品交易量的对应关系。趋势倾向为上涨的概率越高，则买入的交易量越大，理财产品等级越高，则买入的交易量越大。当获取到目标理财产品的趋势倾向和理财产品等级，可以根据趋势倾向与交易量的对应关系和/或理财产品等级与理财产品交易量的对应关系获取对应的理财产品交易量。
需要说明的是，上述的第一理财产品、第二理财产品以及目标理财产品的数量可以为一个或多个，第一理财产品、第二理财产品以及目标理财产品可以完全不同，也可以完全相同，或者部分相同。例如，可以获取几千个股票的信息进行模型训练，然后利用该模型预测A股票、B股票。也可以仅用A股票的信息进行模型训练，然后利用该模型预测A股票。
本申请实施例对目标理财产品的获取方式不进行限定，例如可以根据用户的可以承担的风险类型进行获取，也可以随机获取。也可以根据用户对某一行业的选择选取某一行业理财产品作为目标理财产品。
上述步骤S302以及S304可以同时进行，也可以先后进行。具体可以根据实际需要设置S302以及S304的执行顺序，资产策略配置装置接收并按照该执行顺序执行步骤S302以及S304。例如，为减少目标理财产品的计算数量，可以先获取目标理财产品的趋势倾向，当趋势倾向为上涨时，再获取该目标理财产品的等级。或者，先获取目标理财产品的等级，当等级高于预设等级时，即为绩优股时，再获取目标理财产品的趋势倾向。
在一个实施例中，步骤S304即根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向的步骤包括：当目标理财产品的理财产品等级大于或等于预设等级时，根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向。步骤S306即根据理财产品等级以及趋势倾向获取资产配置策略的步骤包括：当目标理财产品的理财产品等级大于或等于预设等级，且目标理财产品的趋势倾向为上涨时，获取买入目标理财产品的资产配置策略。
以理财产品等级为优、中、差三级为例，若预设等级为中，则当根据理财产品等级模型获取的理财产品等级为中或者优时，将目标理财产品的当前指标状态输入到预设的理财产品趋势模型中，获取目标理财产品的趋势倾向，若趋势倾向为上涨，则获取买入该目标理财产品的资产配置策略。若获取目标理财产品的理财产品等级为差，则不进行理财产品趋势倾向的预测。
上述的资产配置策略获取方法，理财产品等级模型是根据第一理财产品的历史属性信息以及历史属性信息对应的理财产品等级进行模型预训练得到的，理财产品趋势模型是根据第二理财产品的历史指标状态以及历史时刻对应的历史趋势进行模型预训练得到的，由于理财产品的属性信息以及指标状态往往分别蕴含着理财产品的等级以及理财产品后续涨势的规律，故利用历史的理财产品属性信息以及指标状态信息进行模型训练可以准确找出这些数据的规律。因此，当需要获取资产配置策略时，可以根据目标理财产品的当前属性信息以及 预设的理财产品等级模型预测理财产品的理财产品等级，根据目标理财产品的当前指标状态以及预设的理财产品趋势模型预测理财产品的趋势倾向，然后根据目标理财产品的理财产品等级以及理财产品的趋势倾向获取对应的资产配置策略，节约了用户查看大量信息的时间，且获取的资产配置策略准确率高。
在一个实施例中，如图4所示，资产策略配置方法还可以包括以下步骤：
S402，获取第二理财产品在历史时刻之后预设时间的历史价格时间序列，历史价格时间序列包括n+1个价格。
历史价格时间序列为历史时刻以及历史时刻之后预设时间的价格按照时间发展顺序排序组成的序列。预设时间可以根据实际要求进行设置。历史时刻之后预设时间包括历史时刻。每个历史价格之间的时间间隔是相同的。例如可以为连续7天，每天的收盘价。又例如，历史时刻为2016年5月1日，预设时间为7天，则历史价格时间序列为5月1日至5月8日每一天的价格组成的序列，共8个价格。以历史时刻的价格为P0，即起初价格为P0，最终价格为Pn，则历史价格序列可以表示为{P0、P1、P2、......Pn}。
S404，根据历史价格时间序列计算得到趋势判断参数，趋势判断参数包括价格标准分数，或者趋势判断参数包括价格标准分数以及变化效用系数。
其中，价格标准分数为历史价格时间序列中第n+1个价格减去历史价格时间序列平均值的差再除以历史价格时间序列的标准差，变化效用系数为历史价格序列中第t个价格与第一个价格的差的绝对值再除以第一个价格到第t个价格之间相邻价格差的绝对值的和，t等于n或者n+1，n为正整数。可以理解，第n+1个价格指历史价格序列中最终的价格，第一个价格指历史价格序列中最初的价格，以公式表示，价格标准分数A的公式可如公式(1)所示，以t等于n为例，变化效用系数B的公式可如公式(2)所示。
Figure PCTCN2017103954-appb-000001
Figure PCTCN2017103954-appb-000002
其中，P表示历史价格序列，Pn表示第n+1个价格，即预设时间中的最终价格，P0表示第一个价格，即历史时刻的价格也就是起始价格，Pi+1为历史价格序列中第i+2个价格，Pi为历史价格序列中第i+1个价格，其中i大于等于0，小于等于n-1。E(P)为历史价格序列的均值，σ(P)为历史价格序列的标准差。
由公式(1)可知，将第n+1个价格如此变换，可以得到最终价格的值在历史价格序列中的大小位置，若A大于0，表示最终价格在历史价格序列的平均数之上，A小于0，则表示最终价格在历史价格序列的平均数之下。
以n＝4为例进行解释，上述公式(2)的意义如下：设c1、c2、c3分别为P1与P0的差、P2与P1的差、P3与P2的差，则P1、P2、P3可分别表示为P0+c1、P0+c1+c2、P0+c1+c2+c2，故公式2可以化为公式(3)，由公式3可知，B小于等于1且大于等于0，若等于1，说明在该段时间内，理财产品是一直上升或者一直下降的。等于其他值时即有波动，故可以根据B的大小判断理财产品的波动情况。
Figure PCTCN2017103954-appb-000003
S406，根据趋势判断参数获取历史时刻对应的历史趋势。
历史时刻对应的历史趋势指在历史时刻接下来的一段时间第二理财产品相对于历史时刻的涨跌趋势。当趋势判断参数包括价格标准分数时，可以设置当价格标准分数大于预设阈值时则表示历史时刻对应的历史趋势为上涨，小于预设阈值则表示历史时刻对应的历史趋势为下降。或者当趋势判断参数包括价格标准分数以及变化效用系数时，可以结合价格标准分数以及变化效用系数确定历史趋势。
在一个实施例中，为了使得到的理财产品趋势模型更好的预测理财产品趋势，因此，需要选取趋势明显的历史数据进行模型训练，故根据趋势判断参数 获取历史时刻对应的历史趋势的步骤包括：当价格标准分数大于第一预设阈值，且变化效用系数大于第二预设阈值时，得到历史时刻对应的历史趋势为上涨；当价格标准分数的绝对值小于第一预设阈值，且变化效用系数小于第二预设阈值时，得到历史时刻对应的历史趋势为震荡；当价格标准分数的小于第一预设阈值，且变化效用系数大于第二预设阈值时，得到历史时刻对应的历史趋势为下降；其中，第一预设阈值以及第二预设阈值大于0小于等于1。
例如，当第一预设阈值为0.1，第二预设阈值为0.4时，若某一个第二理财产品的价格标准分数为0.5，变化效用系数为0.6，则可得到历史趋势为上涨。
在一些实施例中，当第二预设阈值为0.5时，获取的趋势样本模型预测效果好。
S408，将第二理财产品在历史时刻的历史指标状态以及历史时刻对应的历史趋势组成训练数据，并根据训练数据进行模型训练，得到理财产品趋势模型。
第二理财产品的数量可以为一个也可以为多个，训练数据的数量具体可以根据实际获取，例如可以为几百个也可以为几万个。每个训练数据包括了一个第二理财产品在一个历史时刻下的历史指标状态以及对应的历史趋势。历史指标状态可以为一个或多个。例如，以历史时刻2016年6月16日12时55分为例，则在该历史时刻下，历史指标状态可以包括5日移动平均值大于10日移动平均值、DEA值大于0、DEA值从小往上超过DIF值等，对应的历史趋势为上涨。在获取到训练数据后，根据该训练数据进行机器学习，获得机器学习训练得到的模型参数，得到理财产品趋势模型。
机器学习的模型可以为支持向量机(Support Vector Machine，SVM)分类器模型，神经网络(Artificial Neural Network，ANN)分类器模型，逻辑回归算法(logistic Regression，LR)分类器模型和隐马尔可夫模型(Hidden Markov Model，HMM)等各种进行分类的模型。采用的核函数可以根据实际要求进行设置，例如，在一个实施例中，可以采用支持向量机进行有监督的机器学习，核函数可以采用多项式函数。
在一个实施例中，如图5所示，资产策略配置方法还可以包括以下步骤：
S502，获取由多个第一训练样本构成的第一训练样本集，第一训练样本包 括第一理财产品在历史时刻的多个历史属性信息以及历史属性对应的理财产品等级。
第一理财产品的数量可以为一个或多个，第一训练样本集中第一训练样本数量具体可以根据实际获取，例如可以为几百个也可以为几万个。每个第一训练样本包括了多个历史属性信息以及对应的理财产品等级。例如历史属性信息可以包括市盈率为5％、市现率为10％、换手率为20％等，对应的理财产品等级为绩优股。理财产品等级可以人为标注，也可以通过获取理财产品的数据，进而根据相关量化公式得到。
在一些实施例中，可以通过索提诺比率对第一理财产品的等级进行判断，索提诺比率指每承担一单位的下行波动所能获得的单位收益，数值越大，说明承担同样下跌风险的情况下，可以获得更多的超额收益或者基准上的回报。例如，可以设置当索提诺比例大于第三预设阈值时，为绩优股，小于一定第四预设阈值时，为非绩优股。第三预设阈值以及第四预设阈值可以根据具体要求进行设定。
S504，根据第一训练样本集进行模型训练，得到理财产品等级模型。
在获取到第一训练样本集后，根据该第一样本集进行机器学习，获得机器学习训练得到的模型参数，得到理财产品等级模型。
机器学习的模型可以为支持向量机(Support Vector Machine，SVM)分类器模型，神经网络(Artificial Neural Network，ANN)分类器模型，逻辑回归算法(logistic Regression，LR)分类器模型和隐马尔可夫模型(Hidden Markov Model，HMM)等各种进行分类的模型。采用的核函数可以根据实际要求进行设置，例如，在一个实施例中，可以采用支持向量机进行有监督的机器学习，核函数可以采用多项式函数。
如图6所示，在一个实施例中，提供了一种资产配置策略获取装置，该资产配置策略获取装置可以集成于上述的计算机设备120中，具体可以包括等级获取单元602、趋势倾向获取单元604以及策略获取单元606。
等级获取单元602，用于根据目标理财产品的当前属性信息以及预设的理财 产品等级模型获取目标理财产品的理财产品等级。
预设的理财产品等级模型是根据第一理财产品的历史属性信息以及历史属性信息对应的理财产品等级进行模型预训练得到的。
趋势倾向获取单元604，用于根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向。
预设的理财产品趋势模型是根据第二理财产品在历史时刻的历史指标状态以及历史时刻对应的历史趋势进行模型预训练得到的。
策略获取单元606，用于根据目标理财产品的理财产品等级以及目标理财产品的趋势倾向获取资产配置策略。
在一个实施例中，趋势倾向获取单元604用于：当目标理财产品的理财产品等级大于或等于预设等级时，根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向。策略获取单元606用于：当目标理财产品的理财产品等级大于或等于预设等级，且目标理财产品的趋势倾向为上涨时，获取买入目标理财产品的资产配置策略。
上述的资产配置策略获取装置，理财产品等级模型是根据第一理财产品的历史属性信息以及历史属性信息对应的理财产品等级进行模型预训练得到的，理财产品趋势模型是根据第二理财产品的历史指标状态以及历史时刻对应的历史趋势进行模型预训练得到的，理财产品的属性信息以及指标状态往往分别蕴含着理财产品的等级以及理财产品后续涨势的规律，故利用历史的理财产品属性信息以及指标状态信息进行模型训练可以准确找出这些数据的规律。因此，当需要获取资产配置策略时，可以根据目标理财产品的当前属性信息以及预设的理财产品等级模型预测理财产品的理财产品等级，根据目标理财产品的当前指标状态以及预设的理财产品趋势模型预测理财产品的趋势倾向，然后根据目标理财产品的理财产品等级以及理财产品的趋势倾向获取对应的资产配置策略，节约了用户查看大量信息的时间，且准确率高。
在一个实施例中，如图7所示，资产策略配置装置还可以包括价格时间序列获取单元702、趋势判断参数计算单元704、历史趋势获取单元706以及第二 模型训练单元708：
价格时间序列获取单元702，用于获取第二理财产品在历史时刻之后预设时间的历史价格时间序列，历史价格时间序列包括n+1个价格。
趋势判断参数计算单元704，用于获取历史价格时间序列的趋势判断参数，趋势判断参数包括价格标准分数，或者趋势判断参数包括价格标准分数以及变化效用系数。
历史趋势获取单元706，用于根据趋势判断参数以及预设的判断规则获取历史趋势。
第二模型训练单元708，用于将第二理财产品在历史时刻的历史指标状态以及历史时刻对应的历史趋势组成训练数据，并根据训练数据进行模型训练，得到理财产品趋势模型。
在一个实施例中，如图8所示，资产策略配置装置还可以包括第一样本集获取单元802以及第一模型训练单元804：
第一样本集获取单元802，用于获取由多个第一训练样本构成的第一训练样本集，第一训练样本包括第一理财产品在历史时刻的多个历史属性信息以及历史属性对应的理财产品等级；
第一模型训练单元804，用于根据第一训练样本集进行模型训练，得到理财产品等级模型。
上述资产策略配置装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。其中，网络接口可以是以太网卡或无线网卡等。上述各模块可以硬件形式内嵌于或独立于服务器中的处理器中，也可以以软件形式存储于服务器中的存储器中，以便于处理器调用执行以上各个模块对应的操作。该处理器可以为中央处理单元(CPU)、微处理器、单片机等。
在一个实施例中，本申请提供的资产配置策略获取装置可以实现为一种计算机可读指令的形式，计算机可读指令可在如图2所示的计算机设备上运行，计算机设备的非易失性存储介质可存储组成该资产配置策略获取装置的各个程序模块，比如图6中的等级获取单元602、趋势倾向获取单元604以及策略获取 单元606。各个程序模块中包括计算机可读指令，计算机可读指令用于使计算机设备执行本说明书中描述的本申请各个实施例的资产配置策略获取方法中的步骤。例如，计算机设备可以通过如图6所示的投等级获取单元602根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取目标理财产品的理财产品等级，通过趋势倾向获取单元604根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向，通过策略获取单元606根据目标理财产品的理财产品等级以及目标理财产品的趋势倾向获取资产配置策略。
在一个实施例中，提出了一种计算机设备，计算机设备的内部结构可对应于如图2所示的结构，即该计算机设备既可以是服务器也可以是终端，计算机设备包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机可读指令，处理器执行计算机可读指令时实现以下步骤：根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取目标理财产品的理财产品等级，理财产品等级模型根据第一理财产品的历史属性信息以及历史属性信息对应的理财产品等级进行模型预训练得到；根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向，理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及历史时刻对应的历史趋势进行模型预训练得到；根据目标理财产品的理财产品等级以及目标理财产品的趋势倾向获取资产配置策略。
在一个实施例中，处理器执行计算机可读指令时还执行以下步骤：获取第二理财产品在历史时刻之后预设时间的历史价格时间序列，历史价格时间序列包括n+1个价格；根据历史价格时间序列计算得到趋势判断参数，趋势判断参数包括价格标准分数，或者趋势判断参数包括价格标准分数以及变化效用系数；根据趋势判断参数获取历史时刻对应的历史趋势；将第二理财产品在历史时刻的历史指标状态以及历史时刻对应的历史趋势组成训练数据，并根据训练数据进行模型训练，得到理财产品趋势模型；其中，价格标准分数包括历史价格时间序列中第n+1个价格减去历史价格时间序列平均值的差再除以历史价格时间序列的标准差，变化效用系数为历史价格序列中第t个价格与第一个价格的差的 绝对值再除以第一个价格到第t个价格之间相邻价格差的绝对值的和，t等于n或者n+1，n为正整数。
在一个实施例中，根据趋势判断参数获取历史时刻对应的历史趋势，包括：当价格标准分数大于第一预设阈值，且变化效用系数大于第二预设阈值时，得到历史时刻对应的历史趋势为上涨；当价格标准分数的绝对值小于第一预设阈值，且效用系数小于第二预设阈值时，得到历史时刻对应的历史趋势为震荡；当价格标准分数的小于第一预设阈值，且效用系数大于第二预设阈值时，得到历史时刻对应的历史趋势为下降；其中，第一预设阈值以及第二预设阈值大于0且小于等于1。
在一个实施例中，根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向，包括：当目标理财产品的理财产品等级大于或等于预设等级时，根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向；根据理财产品等级以及趋势倾向获取资产配置策略的步骤包括：当目标理财产品的理财产品等级大于或等于预设等级，且目标理财产品的趋势倾向为上涨时，获取买入目标理财产品的资产配置策略。
在一个实施例中，根据目标理财产品的理财产品等级以及目标理财产品的趋势倾向获取资产配置策略，还包括：根据目标理财产品的趋势倾向和/或目标理财产品的理财产品等级确定目标理财产品对应的交易量。
在一个实施例中，提出了一种存储有计算机可读指令的存储介质，该计算机可读指令被一个或多个处理器执行时，使得一个或多个处理器执行以下步骤：根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取目标理财产品的理财产品等级，理财产品等级模型根据第一理财产品的历史属性信息以及历史属性信息对应的理财产品等级进行模型预训练得到；根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向，理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及历史时刻对应的历史趋势进行模型预训练得到；根据目标理财产品的理财产品等级以及 目标理财产品的趋势倾向获取资产配置策略。
在一个实施例中，处理器执行计算机可读指令时还执行以下步骤：获取第二理财产品在历史时刻之后预设时间的历史价格时间序列，历史价格时间序列包括n+1个价格；根据历史价格时间序列计算得到趋势判断参数，趋势判断参数包括价格标准分数，或者趋势判断参数包括价格标准分数以及变化效用系数；根据趋势判断参数获取历史时刻对应的历史趋势；将第二理财产品在历史时刻的历史指标状态以及历史时刻对应的历史趋势组成训练数据，并根据训练数据进行模型训练，得到理财产品趋势模型；其中，价格标准分数包括历史价格时间序列中第n+1个价格减去历史价格时间序列平均值的差再除以历史价格时间序列的标准差，变化效用系数为历史价格序列中第t个价格与第一个价格的差的绝对值再除以第一个价格到第t个价格之间相邻价格差的绝对值的和，t等于n或者n+1，n为正整数。
在一个实施例中，根据趋势判断参数获取历史时刻对应的历史趋势，包括：当价格标准分数大于第一预设阈值，且变化效用系数大于第二预设阈值时，得到历史时刻对应的历史趋势为上涨；当价格标准分数的绝对值小于第一预设阈值，且效用系数小于第二预设阈值时，得到历史时刻对应的历史趋势为震荡；当价格标准分数的小于第一预设阈值，且效用系数大于第二预设阈值时，得到历史时刻对应的历史趋势为下降；其中，第一预设阈值以及第二预设阈值大于0且小于等于1。
在一个实施例中，根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向，包括：当目标理财产品的理财产品等级大于或等于预设等级时，根据目标理财产品的当前指标状态以及预设的理财产品趋势模型获取目标理财产品的趋势倾向；根据理财产品等级以及趋势倾向获取资产配置策略的步骤包括：当目标理财产品的理财产品等级大于或等于预设等级，且目标理财产品的趋势倾向为上涨时，获取买入目标理财产品的资产配置策略。
在一个实施例中，根据目标理财产品的理财产品等级以及目标理财产品的趋势倾向获取资产配置策略，还包括：根据目标理财产品的趋势倾向和/或目标 理财产品的理财产品等级确定目标理财产品对应的交易量。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机可读指令来指令相关的硬件来完成，该计算机可读指令可存储于一计算机可读取存储介质中，该程序在执行时，可包括如上述各方法的实施例的流程。其中，前述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory，ROM)等非易失性存储介质等。
以上所述实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。
以上所述实施例仅表达了本申请的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对本申请专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本申请构思的前提下，还可以做出若干变形和改进，这些都属于本申请的保护范围。因此，本申请专利的保护范围应以所附权利要求为准。

Claims (20)

  1. 一种资产配置策略获取方法，包括：
    根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取所述目标理财产品的理财产品等级，所述理财产品等级模型根据第一理财产品的历史属性信息以及所述历史属性信息对应的理财产品等级进行模型预训练得到；
    根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，所述理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及所述历史时刻对应的历史趋势进行模型预训练得到；
    根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略。
  2. 根据权利要求1所述的方法，其特征在于，所述方法还包括：
    获取所述第二理财产品在所述历史时刻之后预设时间的历史价格时间序列，所述历史价格时间序列包括n+1个价格；
    根据所述历史价格时间序列计算得到趋势判断参数，所述趋势判断参数包括价格标准分数，或者所述趋势判断参数包括所述价格标准分数以及变化效用系数；
    根据所述趋势判断参数获取所述历史时刻对应的历史趋势；
    将所述第二理财产品在所述历史时刻的历史指标状态以及所述历史时刻对应的历史趋势组成训练数据，并根据所述训练数据进行模型训练，得到所述理财产品趋势模型；
    其中，所述价格标准分数包括所述历史价格时间序列中第n+1个价格减去历史价格时间序列平均值的差再除以历史价格时间序列的标准差，所述变化效用系数为所述历史价格序列中第t个价格与第一个价格的差的绝对值再除以所述第一个价格到第t个价格之间相邻价格差的绝对值的和，t等于n或者n+1，所述n为正整数。
  3. 根据权利要求2所述的方法，其特征在于，所述根据所述趋势判断参数 获取所述历史时刻对应的历史趋势的步骤包括：
    当所述价格标准分数大于第一预设阈值，且所述变化效用系数大于第二预设阈值时，得到所述历史时刻对应的历史趋势为上涨；
    当所述价格标准分数的绝对值小于所述第一预设阈值，且所述效用系数小于所述第二预设阈值时，得到所述历史时刻对应的历史趋势为震荡；
    当所述价格标准分数的小于所述第一预设阈值，且所述效用系数大于第二预设阈值时，得到所述历史时刻对应的历史趋势为下降；
    其中，所述第一预设阈值以及第二预设阈值大于0且小于等于1。
  4. 根据权利要求1所述的方法，其特征在于，所述根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向的步骤包括：
    当所述目标理财产品的理财产品等级大于或等于预设等级时，根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向；
    所述根据所述理财产品等级以及所述趋势倾向获取资产配置策略的步骤包括：
    当所述目标理财产品的理财产品等级大于或等于预设等级，且所述目标理财产品的趋势倾向为上涨时，获取买入所述目标理财产品的资产配置策略。
  5. 根据权利要求4所述的方法，其特征在于，所述根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略的步骤还包括：
    根据所述目标理财产品的趋势倾向和/或所述目标理财产品的理财产品等级确定所述目标理财产品对应的交易量。
  6. 一种资产配置策略获取装置，包括：
    等级获取单元，用于根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取所述目标理财产品的理财产品等级，所述理财产品等级模型根据第一理财产品的历史属性信息以及所述历史属性信息对应的理财产品等级进行模型预训练得到；
    趋势倾向获取单元，用于根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，所述理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及所述历史时刻对应的历史趋势进行模型预训练得到；
    策略获取单元，用于根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略。
  7. 根据权利要求6所述的装置，其特征在于，所述装置还包括：
    价格时间序列获取单元，用于获取所述第二理财产品在所述历史时刻之后预设时间的历史价格时间序列，所述历史价格时间序列包括n+1个价格；
    趋势判断参数计算单元，用于根据所述历史价格时间序列计算得到趋势判断参数，所述趋势判断参数包括价格标准分数，或者所述趋势判断参数包括所述价格标准分数以及变化效用系数；
    历史趋势获取单元，用于根据所述趋势判断参数获取所述历史时刻对应的历史趋势；
    第二模型训练单元，用于将所述第二理财产品在所述历史时刻的历史指标状态以及所述历史时刻对应的历史趋势组成训练数据，并根据所述训练数据进行模型训练，得到所述理财产品趋势模型；
    其中，所述价格标准分数包括所述历史价格时间序列中第n+1个价格减去历史价格时间序列平均值的差再除以历史价格时间序列的标准差，所述变化效用系数为所述历史价格序列中第t个价格与第一个价格的差的绝对值再除以所述第一个价格到第t个价格之间相邻价格差的绝对值的和，t等于n或者n+1，所述n为正整数。
  8. 根据权利要求7所述的装置，其特征在于，所述历史趋势获取单元具体用于：
    当所述价格标准分数大于第一预设阈值，且所述变化效用系数大于第二预设阈值时，得到所述历史时刻对应的历史趋势为上涨；
    当所述价格标准分数的绝对值小于所述第一预设阈值，且所述效用系数小于所述第二预设阈值时，得到所述历史时刻对应的历史趋势为震荡；
    当所述价格标准分数的小于所述第一预设阈值，且所述效用系数大于第二预设阈值时，得到所述历史时刻对应的历史趋势为下降；
    其中，所述第一预设阈值以及第二预设阈值大于0且小于等于1。
  9. 根据权利要求6所述的装置，其特征在于，所述趋势倾向获取单元具体用于：
    当所述目标理财产品的理财产品等级大于或等于预设等级时，根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向；
    所述策略获取单元用于：
    当所述目标理财产品的理财产品等级大于或等于预设等级，且所述目标理财产品的趋势倾向为上涨时，获取买入所述目标理财产品的资产配置策略。
  10. 根据权利要求9所述的装置，其特征在于，所述策略获取单元还用于：
    根据所述目标理财产品的趋势倾向和/或所述目标理财产品的理财产品等级确定所述目标理财产品对应的交易量。
  11. 一种计算机设备，包括存储器和处理器，所述存储器中存储有计算机可读指令，所述计算机可读指令被所述处理器执行时，使得所述处理器执行以下步骤：
    根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取所述目标理财产品的理财产品等级，所述理财产品等级模型根据第一理财产品的历史属性信息以及所述历史属性信息对应的理财产品等级进行模型预训练得到；
    根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，所述理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及所述历史时刻对应的历史趋势进行模型预训练得到；
    根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略。
  12. 根据权利要求11所述的计算机设备，其特征在于，所述处理器还执行 以下步骤：
    获取所述第二理财产品在所述历史时刻之后预设时间的历史价格时间序列，所述历史价格时间序列包括n+1个价格；
    根据所述历史价格时间序列计算得到趋势判断参数，所述趋势判断参数包括价格标准分数，或者所述趋势判断参数包括所述价格标准分数以及变化效用系数；
    根据所述趋势判断参数获取所述历史时刻对应的历史趋势；
    将所述第二理财产品在所述历史时刻的历史指标状态以及所述历史时刻对应的历史趋势组成训练数据，并根据所述训练数据进行模型训练，得到所述理财产品趋势模型；
    其中，所述价格标准分数包括所述历史价格时间序列中第n+1个价格减去历史价格时间序列平均值的差再除以历史价格时间序列的标准差，所述变化效用系数为所述历史价格序列中第t个价格与第一个价格的差的绝对值再除以所述第一个价格到第t个价格之间相邻价格差的绝对值的和，t等于n或者n+1，所述n为正整数。
  13. 根据权利要求12所述的计算机设备，其特征在于，所述处理器所执行的所述根据所述趋势判断参数获取所述历史时刻对应的历史趋势，包括：
    当所述价格标准分数大于第一预设阈值，且所述变化效用系数大于第二预设阈值时，得到所述历史时刻对应的历史趋势为上涨；
    当所述价格标准分数的绝对值小于所述第一预设阈值，且所述效用系数小于所述第二预设阈值时，得到所述历史时刻对应的历史趋势为震荡；
    当所述价格标准分数的小于所述第一预设阈值，且所述效用系数大于第二预设阈值时，得到所述历史时刻对应的历史趋势为下降；
    其中，所述第一预设阈值以及第二预设阈值大于0且小于等于1。
  14. 根据权利要求11所述的计算机设备，其特征在于，所述处理器所执行的所述根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，包括：
    当所述目标理财产品的理财产品等级大于或等于预设等级时，根据所述目 标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向；
    所述处理器所执行的所述根据所述理财产品等级以及所述趋势倾向获取资产配置策略，包括：
    当所述目标理财产品的理财产品等级大于或等于预设等级，且所述目标理财产品的趋势倾向为上涨时，获取买入所述目标理财产品的资产配置策略。
  15. 根据权利要求14所述的计算机设备，其特征在于，所述处理器所执行的所述根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略，还包括：
    根据所述目标理财产品的趋势倾向和/或所述目标理财产品的理财产品等级确定所述目标理财产品对应的交易量。
  16. 一个或多个存储有计算机可读指令的非易失性可读存储介质，所述计算机可读指令被一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：
    根据目标理财产品的当前属性信息以及预设的理财产品等级模型获取所述目标理财产品的理财产品等级，所述理财产品等级模型根据第一理财产品的历史属性信息以及所述历史属性信息对应的理财产品等级进行模型预训练得到；
    根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，所述理财产品趋势模型根据第二理财产品在历史时刻的历史指标状态以及所述历史时刻对应的历史趋势进行模型预训练得到；
    根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略。
  17. 根据权利要求16所述的存储介质，其特征在于，所述处理器还执行以下步骤：
    获取所述第二理财产品在所述历史时刻之后预设时间的历史价格时间序列，所述历史价格时间序列包括n+1个价格；
    根据所述历史价格时间序列计算得到趋势判断参数，所述趋势判断参数包括价格标准分数，或者所述趋势判断参数包括所述价格标准分数以及变化效用系数；
    根据所述趋势判断参数获取所述历史时刻对应的历史趋势；
    将所述第二理财产品在所述历史时刻的历史指标状态以及所述历史时刻对应的历史趋势组成训练数据，并根据所述训练数据进行模型训练，得到所述理财产品趋势模型；
    其中，所述价格标准分数包括所述历史价格时间序列中第n+1个价格减去历史价格时间序列平均值的差再除以历史价格时间序列的标准差，所述变化效用系数为所述历史价格序列中第t个价格与第一个价格的差的绝对值再除以所述第一个价格到第t个价格之间相邻价格差的绝对值的和，t等于n或者n+1，所述n为正整数。
  18. 根据权利要求17所述的存储介质，其特征在于，所述处理器所执行的所述根据所述趋势判断参数获取所述历史时刻对应的历史趋势，包括：
    当所述价格标准分数大于第一预设阈值，且所述变化效用系数大于第二预设阈值时，得到所述历史时刻对应的历史趋势为上涨；
    当所述价格标准分数的绝对值小于所述第一预设阈值，且所述效用系数小于所述第二预设阈值时，得到所述历史时刻对应的历史趋势为震荡；
    当所述价格标准分数的小于所述第一预设阈值，且所述效用系数大于第二预设阈值时，得到所述历史时刻对应的历史趋势为下降；
    其中，所述第一预设阈值以及第二预设阈值大于0且小于等于1。
  19. 根据权利要求16所述的存储介质，其特征在于，所述处理器所执行的所述根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向，包括：
    当所述目标理财产品的理财产品等级大于或等于预设等级时，根据所述目标理财产品的当前指标状态以及预设的理财产品趋势模型获取所述目标理财产品的趋势倾向；
    所述根据所述理财产品等级以及所述趋势倾向获取资产配置策略的步骤包 括：
    当所述目标理财产品的理财产品等级大于或等于预设等级，且所述目标理财产品的趋势倾向为上涨时，获取买入所述目标理财产品的资产配置策略。
  20. 根据权利要求19所述的存储介质，其特征在于，所述处理器所执行的所述根据所述目标理财产品的理财产品等级以及所述目标理财产品的趋势倾向获取资产配置策略，还包括：
    根据所述目标理财产品的趋势倾向和/或所述目标理财产品的理财产品等级确定所述目标理财产品对应的交易量。
PCT/CN2017/103954 2017-07-25 2017-09-28 资产配置策略获取方法、装置、计算机设备和存储介质 Ceased WO2019019346A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
SG11201913398QA SG11201913398QA (en) 2017-07-25 2017-09-28 Asset allocation strategy acquisition method and apparatus, computer device, and storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710614428.9 2017-07-25
CN201710614428.9A CN107679987A (zh) 2017-07-25 2017-07-25 资产配置策略获取方法、装置、计算机设备和存储介质

Publications (1)

Publication Number Publication Date
WO2019019346A1 true WO2019019346A1 (zh) 2019-01-31

Family

ID=61133548

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/103954 Ceased WO2019019346A1 (zh) 2017-07-25 2017-09-28 资产配置策略获取方法、装置、计算机设备和存储介质

Country Status (3)

Country Link
CN (1) CN107679987A (zh)
SG (1) SG11201913398QA (zh)
WO (1) WO2019019346A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113570207A (zh) * 2021-07-09 2021-10-29 北京淇瑀信息科技有限公司 用户策略分配方法、装置及电子设备
CN115827821A (zh) * 2022-11-11 2023-03-21 深圳市今古科技有限公司 一种基于资讯信息的判断策略生成方法及系统
CN116050816A (zh) * 2022-12-19 2023-05-02 北京爱笔科技有限公司 行为调整策略的确定方法、装置、计算机设备和存储介质
CN118428909A (zh) * 2024-07-04 2024-08-02 成都中智游科技有限公司 基于大数据的文旅建设用智慧管理系统

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109035028B (zh) * 2018-06-29 2023-08-22 平安科技（深圳）有限公司 智能投顾策略生成方法及装置、电子设备、存储介质
CN109523342A (zh) * 2018-10-12 2019-03-26 平安科技（深圳）有限公司 服务策略生成方法及装置、电子设备、存储介质
CN109783876B (zh) * 2018-12-19 2024-02-06 平安科技（深圳）有限公司 时间序列模型建立方法、装置、计算机设备和存储介质
CN111860855B (zh) * 2019-12-18 2023-12-05 北京嘀嘀无限科技发展有限公司 一种行为引导资源投放策略生成方法及装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110137821A1 (en) * 2009-12-07 2011-06-09 Predictive Technologies Group, Llc Calculating predictive technical indicators
CN106022522A (zh) * 2016-05-20 2016-10-12 南京大学 一种基于互联网公开的大数据预测股票的方法及系统
CN106203731A (zh) * 2016-07-25 2016-12-07 上海电机学院 一种基于2‑维流预测的股票趋势预测方法
CN106600404A (zh) * 2016-11-08 2017-04-26 南京财经大学 一种预测方法及装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110137821A1 (en) * 2009-12-07 2011-06-09 Predictive Technologies Group, Llc Calculating predictive technical indicators
CN106022522A (zh) * 2016-05-20 2016-10-12 南京大学 一种基于互联网公开的大数据预测股票的方法及系统
CN106203731A (zh) * 2016-07-25 2016-12-07 上海电机学院 一种基于2‑维流预测的股票趋势预测方法
CN106600404A (zh) * 2016-11-08 2017-04-26 南京财经大学 一种预测方法及装置

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113570207A (zh) * 2021-07-09 2021-10-29 北京淇瑀信息科技有限公司 用户策略分配方法、装置及电子设备
CN113570207B (zh) * 2021-07-09 2024-04-26 北京淇瑀信息科技有限公司 用户策略分配方法、装置及电子设备
CN115827821A (zh) * 2022-11-11 2023-03-21 深圳市今古科技有限公司 一种基于资讯信息的判断策略生成方法及系统
CN115827821B (zh) * 2022-11-11 2024-01-09 深圳市今古科技有限公司 一种基于资讯信息的判断策略生成方法及系统
CN116050816A (zh) * 2022-12-19 2023-05-02 北京爱笔科技有限公司 行为调整策略的确定方法、装置、计算机设备和存储介质
CN118428909A (zh) * 2024-07-04 2024-08-02 成都中智游科技有限公司 基于大数据的文旅建设用智慧管理系统

Also Published As

Publication number Publication date
CN107679987A (zh) 2018-02-09
SG11201913398QA (en) 2020-01-30

Similar Documents

Publication Publication Date Title
WO2019019346A1 (zh) 资产配置策略获取方法、装置、计算机设备和存储介质
AU2021232839B2 (en) Updating Attribute Data Structures to Indicate Trends in Attribute Data Provided to Automated Modelling Systems
US11586941B2 (en) Recommendation method and apparatus
TWI726341B (zh) 樣本屬性評估模型訓練方法、裝置、伺服器及儲存媒體
US11568480B2 (en) Artificial intelligence derived anonymous marketplace
US11113704B2 (en) Systems and methods for interactive annuity product services using machine learning modeling
US11544627B1 (en) Machine learning-based methods and systems for modeling user-specific, activity specific engagement predicting scores
CN107798607A (zh) 资产配置策略获取方法、装置、计算机设备和存储介质
CN111340244B (zh) 预测方法、训练方法、装置、服务器及介质
US20170185922A1 (en) Hierarchical Capital Allocation Using Clustered Machine Learning
CN110633989B (zh) 一种风险行为生成模型的确定方法及装置
US20170364825A1 (en) Adaptive augmented decision engine
CN106462795A (zh) 用于针对金融市场中的大数据交易将资本分配到交易策略的系统及方法
CN112966189A (zh) 一种基金产品推荐系统
US11367135B2 (en) Systems and methods for intelligently optimizing a queue of actions in an interface using machine learning
CN114581249A (zh) 基于投资风险承受能力评估的金融产品推荐方法及系统
WO2019205384A1 (zh) 电子装置、基于机器学习的股票择时方法及存储介质
CN106600404A (zh) 一种预测方法及装置
CN106897282B (zh) 一种用户群的分类方法和设备
US20160035026A1 (en) Online trading systems and methods
CN113989051B (zh) 产品推荐方法及装置
CN114782201A (zh) 股票推荐方法、装置、计算机设备和存储介质
CN114219315A (zh) 营销效果的评价方法、装置、计算机设备以及存储介质
US20210103990A1 (en) Automated real time mortgage servicing and whole loan valuation
Hadi et al. Predicting Bitcoin and Ethereum prices using long short-term memory and gated recurrent unit

Legal Events

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

Ref document number: 17919633

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17919633

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 28/05/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 17919633

Country of ref document: EP

Kind code of ref document: A1