WO2017133615A1 - 一种业务参数获取方法及装置 - Google Patents
一种业务参数获取方法及装置 Download PDFInfo
- Publication number
- WO2017133615A1 WO2017133615A1 PCT/CN2017/072593 CN2017072593W WO2017133615A1 WO 2017133615 A1 WO2017133615 A1 WO 2017133615A1 CN 2017072593 W CN2017072593 W CN 2017072593W WO 2017133615 A1 WO2017133615 A1 WO 2017133615A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- parameter
- sample user
- feature
- logistic regression
- regression analysis
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
- H04L41/5061—Network service management, e.g. ensuring proper service fulfilment according to agreements characterised by the interaction between service providers and their network customers, e.g. customer relationship management
- H04L41/5064—Customer relationship management
Definitions
- the present application relates to the field of Internet technologies, and in particular, to a method and an apparatus for acquiring a service parameter.
- Service parameters directly affect whether a service application can be successful.
- a service provider allocates a service to a user, it evaluates whether the service is assigned to the user based on the existing service parameters.
- the service provider has a large number of user service parameter records, and needs to obtain the required target user service parameters.
- the service provider cannot accurately accurately evaluate the required target user's service parameters.
- the embodiment of the present application provides a method and an apparatus for acquiring a service parameter.
- An object of the present application is to provide a method for acquiring a service parameter, the method comprising:
- the logistic regression analysis model is obtained by performing logistic regression analysis and repeated iterative training using characteristic data of a large number of sample users.
- Another object of the present application is to provide a service parameter obtaining apparatus, the apparatus comprising:
- An obtaining unit configured to acquire feature data of a sample user of a service parameter to be predicted
- a processing unit configured to determine that the sample user that meets the preset rule is the target sample user
- a further object of the present application is to provide a service parameter obtaining device, which includes a processor and a memory for storing a program for supporting a data processing device to execute the above method, the processor being configured It is used to execute a program stored in the memory.
- the database processing device can also include a communication interface for the database processing device to communicate with other devices or communication networks.
- the embodiment of the present application provides a computer storage medium for storing computer software instructions used by the service parameter obtaining apparatus, which includes a program designed to execute the foregoing aspect for the service parameter obtaining apparatus.
- FIG. 1 is a flowchart of an embodiment of a method for acquiring a service parameter in an embodiment of the present application
- FIG. 2 is a flowchart of another embodiment of a method for acquiring a service parameter according to an embodiment of the present application
- FIG. 3 is a structural diagram of an embodiment of a service parameter obtaining apparatus according to an embodiment of the present application.
- FIG. 4 is a structural diagram of another embodiment of a service parameter obtaining apparatus according to an embodiment of the present application.
- the logistic regression analysis model is based on a machine learning model with supervised training.
- Supervised learning A training method with training samples and training tags.
- the present application determines the corresponding service parameters by using the characteristic data of the user.
- these service parameters can reflect the integrity of the user in a certain period of time, that is, whether the default situation will occur.
- the application can reflect whether the user can default the business parameters. Is the probability of default, that is, between 0 and 1, if the probability of default of the business parameter tends to 0, it indicates that the probability of default is small, for example, the probability of default is 0.1. On the contrary, if the probability of default is more toward 1, it indicates The probability of default is greater, such as a default probability of 0.9.
- the default prediction and the user default probability in the embodiment of the present application are only different in expression, and the principle is the same.
- the method for obtaining the service parameter is provided by the embodiment of the present application, and the method includes:
- S102 Enter the feature data into a logistic regression analysis model to obtain a feature parameter of the feature data, where the feature parameter is used to determine the service parameter, where the logistic regression analysis model is performed by using feature data of a large number of sample users. Logistic regression analysis and repeated iteration training.
- logistic regression analysis In the logistic regression analysis model, a large number of sample users can be analyzed in advance to obtain the commonly used characteristic parameters of the sample users.
- logistic regression analysis can be used to analyze each characteristic parameter already existing in the model.
- the corresponding value is determined, because the corresponding sample user can have multiple characteristic parameters, and each of the characteristic parameters is different for the sample user.
- the sample user has A characteristic parameter, B characteristic parameter and C characteristic respectively.
- the parameters, the corresponding values can be 0.2, 0.5 and 0.3 respectively.
- the characteristic parameters can be used to determine the business parameters.
- the business parameters here can represent the credit degree of the sample users, and the characteristic parameters are 0, 1.
- the characteristic parameter tends to 1
- the probability of default is small, that is, the credit is very high, and the intermediate value can usually be selected. Dividing, for example, between 0 and 0.5 as the first threshold interval, between 0.5 and 1 The second threshold interval is determined.
- the sample user When the feature parameter of the sample user is within the first threshold interval, the sample user may be determined to have the first service parameter, and when the feature parameter of the sample user is located in the second threshold interval, the sample user may be determined to have the second Business parameters, because the logistic regression analysis model is a numerical value corresponding to the characteristic parameters determined by analyzing a large number of sample users in advance, so that when a business parameter is acquired for a user whose business parameters are to be tested If it is more accurate, it can objectively estimate the default of the sample users.
- the embodiment of the present application provides a method for acquiring a service parameter, where the method includes:
- the preset rule includes: a user whose location is located at the target location, a degree of association with the target sample user reaches a preset association threshold, and the identity information of the sample user meets a preset condition, for example, in progress
- the position of the student can be matched with the geographical location of the major universities in the country.
- the location function of the equipment can be used for the position of the student, and the user should be authorized for the position of the student.
- students who are determined to be sample users students can be extended according to their associated circle of friends to expand more sample users who meet the conditions of the students, so that a large number of samples can be used for determining the students' samples, and the logistic regression analysis can be improved.
- the accuracy of the model can be improved.
- the characteristic parameters may include statistical analysis on the sample user location migration frequency, the contact update frequency, the push frequency of the social application information, etc., which may be obtained through statistics, and then the repeated feature calculations are performed to determine accurate feature parameters and corresponding parameters.
- the value that is, the weight value, for example, the frequency of a person's location migration, the location of the occurrence is not fixed, it can be considered that the user's work or learning state is unstable, when the business is assigned to it, the later progress may not be smooth.
- the weight value of the feature parameter can be increased to reflect the importance. For example, when a loan is made to the user, due to unstable work or learning, there will be a situation in which the payment cannot be repaid on time. Such a user default risk will increase, and then more review will be conducted when the loan is made.
- S205 When the feature parameter is located in a preset first threshold interval, determine that the sample user has the first service parameter, and when the feature parameter is located in a preset second threshold interval, determine the sample user. Having the second service parameter.
- the feature parameter output by the logistic regression analysis model according to the feature data may be a probability value, the range of the feature parameter is between 0 and 1, and the service parameter is divided into two types including the first service parameter and the second service parameter, and the first service
- the parameter can also be set as a good faith user, and the second service parameter can be set as a default user.
- the service parameter can correspond to the user's default possibility, so that the corresponding user can be a good user and a default user, such as a characteristic parameter.
- the sample user has more features of the honest user, and it can be said that the sample user is less likely to default.
- the feature parameter is between 0.5 and 1, the sample user has a default. The user has more features.
- the sample user has a higher probability of default.
- the user can flexibly choose.
- the value of the intermediate value can be closer to 0.
- the first The threshold interval can be set between 0 and 0.2
- the second threshold interval is set between 0.2 and 1, correspondingly, for integrity If the condition of the household is loose, the value of the intermediate value may be closer to 1, for example, 0.7
- the first threshold interval may be set to 0 to 0.7
- the second threshold interval may be set to 0.7 to 1, in short, by the characteristic parameter
- the value of the sample user can determine the business parameters of the sample user, and the sample user's default condition can be predicted.
- N1 parameters are selected from the second parameter by the cluster analysis, and the second parameter is analyzed by the discriminant analysis Select N2 parameters, combine the selected N1 parameters and N2 parameters to obtain the third parameter;
- Probability of defaulting users A y value of 1 indicates a default customer, and a 0 is a good customer.
- ⁇ represents the parameters estimated by the model, namely: ⁇ , ⁇ 1 , ⁇ 2 , ..., ⁇ n
- log(L( ⁇ )) By deriving log(L( ⁇ )), the extremum is obtained, and the iterative function of ⁇ is obtained, which is the estimated parameter of the logistic regression analysis model.
- the actual corresponding estimated parameter of the model variable can be used as the weight value of each feature parameter.
- the premise of the selection of the logistic regression analysis model variable is the derivative variable.
- the object of the analysis may be the user or the account.
- the obtained data may have user basic attribute data, social attribute data, transaction attribute data, stable security. Attribute variables and the like can be derived from the data to obtain new variables for use. The process of creating the derived variables should be understood by those of ordinary skill in the art, and will not be described herein.
- the embodiment of the present application discloses a method for acquiring a service parameter, which first acquires a service parameter to be predicted. Feature data of the sample user, the feature data is input to a logistic regression analysis model to obtain feature parameters of the feature data, and the feature parameters are used to determine the service parameter, wherein the logistic regression analysis model adopts a large number of The characteristic data of the sample user is subjected to logistic regression analysis and iteratively iterative training is obtained. Because the logistic regression analysis model pre-analyzes the values corresponding to the characteristic parameters determined by a large number of sample users, the business parameter acquisition is performed on the user of a service parameter to be tested. The results are more accurate and can objectively estimate the default of the sample users.
- the embodiment of the present application further provides a service parameter obtaining device, where the device includes:
- An obtaining unit 301 configured to acquire feature data of a sample user of a service parameter to be predicted
- the analyzing unit 302 is configured to perform categorization analysis on the feature data by using a logistic regression analysis model to obtain a plurality of feature parameters of the feature data;
- An obtaining unit 301 configured to acquire feature data of a sample user of a service parameter to be predicted
- the processing unit 302 is configured to input the feature data into a logistic regression analysis model to obtain a feature parameter of the feature data, where the feature parameter is used to determine the service parameter, wherein the logistic regression analysis model uses a large number of sample users
- the eigendata is subjected to logistic regression analysis and iterative training is obtained.
- processing unit 302 is further configured to:
- a method for determining the logistic regression analysis model using a plurality of feature data of the target sample user is a method for determining the logistic regression analysis model using a plurality of feature data of the target sample user.
- the service parameter includes a first service parameter and a second service parameter
- the processing unit 302 is further configured to:
- the feature parameter When the feature parameter is located in a preset second threshold interval, it is determined that the sample user has the second service parameter.
- processing unit 302 is further configured to:
- the preset rule includes: the user where the sample user is located at the target location, the degree of association with the target sample user reaches a preset association threshold, and the identity information of the sample user conforms to the preset. condition.
- the embodiment of the present application discloses a service parameter obtaining device, which first obtains feature data of a sample user of a service parameter to be predicted, and uses a logistic regression analysis model to classify the feature data to obtain multiple features of the feature data.
- a parameter determining a value of each of the plurality of characteristic parameters for determining the business parameter, wherein the logistic regression analysis model performs logistic regression analysis using a plurality of sample user characteristic data and repeats Iterative training is obtained, because the logistic regression analysis model pre-calculates the values corresponding to the characteristic parameters determined by a large number of sample users, so that the results of the business parameters acquisition for a user whose business parameters are to be tested are more accurate, and the sample users can be more objective. The default is estimated.
- FIG. 4 is a schematic structural diagram of a service parameter obtaining apparatus 40 according to an embodiment of the present application.
- the service parameter acquisition device 40 includes a processor 410, a memory 450, and an input/output I/O device 430.
- the memory 450 can include read only memory and random access memory, and provides operational instructions and data to the processor 410.
- a portion of the memory 450 may also include non-volatile random access memory (NVRAM).
- NVRAM non-volatile random access memory
- the memory 450 stores the following elements, executable modules or data structures, or a subset thereof, or their extended set:
- the operation instruction can be stored in the operating system
- the feature parameters are used to determine the business parameter, wherein the logistic regression analysis model uses a feature data of a large number of sample users for logistic Regression analysis and iterative training.
- the processor 410 controls the operation of the service parameter obtaining device 40, which may also be referred to as CPU (Central Processing Unit).
- Memory 450 can include read only memory and random access memory and provides instructions and data to processor 410. A portion of the memory 450 may also include non-volatile random access memory (NVRAM).
- the various components of the business parameter acquisition device 40 in the application are coupled together by a bus system 420.
- the bus system 420 may include a power bus, a control bus, a status signal bus, and the like in addition to the data bus. However, for clarity of description, various buses are labeled as bus system 420 in the figure.
- Processor 410 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the foregoing method may be completed by an integrated logic circuit of hardware in the processor 410 or an instruction in a form of software.
- the processor 410 described above may be a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, or discrete hardware. Component.
- DSP digital signal processor
- ASIC application specific integrated circuit
- FPGA off-the-shelf programmable gate array
- the methods, steps, and logical block diagrams disclosed in the embodiments of the present application can be implemented or executed.
- the general purpose processor may be a microprocessor or the processor or any conventional processor or the like.
- the steps of the method disclosed in the embodiments of the present application may be directly implemented by the hardware decoding processor, or may be performed by a combination of hardware and software modules in the decoding processor.
- the software module can be located in a conventional storage medium such as random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, registers, and the like.
- the storage medium is located in the memory 450, and the processor 410 reads the information in the memory 450 and completes the steps of the above method in combination with its hardware.
- the disclosed system, apparatus, and method may be implemented in other manners.
- the device embodiments described above are merely illustrative.
- the division of the unit is only a logical function division.
- there may be another division manner for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed.
- the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.
- the units described as separate components may or may not be physically separate.
- the components displayed for the unit may or may not be physical units, ie may be located in one place, or may be distributed over multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
- each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
- the above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.
- the program may be stored in a computer readable storage medium, and the storage medium may include: Read Only Memory (ROM), Random Access Memory (RAM), disk or optical disk.
- ROM Read Only Memory
- RAM Random Access Memory
Landscapes
- Business, Economics & Management (AREA)
- General Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
本申请实施例公开了一种业务参数获取方法及装置,首先获取待预测业务参数的样本用户的特征数据,将所述特征数据输入到logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数,当所述特征参数位于预设的第一阈值区间时,确定所述样本用户具有所述第一业务参数,当所述特征参数位于预设的第二阈值区间时,确定所述样本用户具有所述第二业务参数,其中所述logistic回归分析模型是采用大量样本用户的特征数据进行logistic回归分析并反复迭代训练得到,因为logistic回归分析模型预先对大量的样本用户进行分析后确定的特征参数对应的数值,这样对一个待测试业务参数的用户进行业务参数获取时候结果比较准确,能够较为客观对样本用户的违约进行预估。
Description
本申请要求于2016年2月3日提交中国专利局、申请号为201610078384.8、发明名称为“一种业务参数获取方法及装置”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。
本申请涉及互联网技术领域,特别涉及一种业务参数获取方法及装置。
当前很多业务与业务参数都是直接相关的,业务参数直接影响到业务申请是否能够成功。业务提供方在为用户分配业务时会根据已有的业务参数来评估是否为该用户分配业务。
但目前,在业务提供方有可以获得大量的用户业务参数记录,需要从中获取到所需要的目标用户的业务参数,目前业务提供方无法准确对所需要的目标用户的业务参数进行准确的评估,导致业务提供方目标用户提供业务存在一定风险。
发明内容
有鉴于此,本申请实施例提供了一种业务参数获取方法及装置。
本申请的一个目的是提供一种业务参数获取方法,所述方法包括:
确定满足预置规则的样本用户为目标样本用户;
利用大量所述目标样本用户的特征数据确定logistic回归分析模型;
获取待预测业务参数的样本用户的特征数据;
将所述特征数据输入到所述logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数;
当所述特征参数位于预设的第一阈值区间时,确定所述样本用户具有所述第一业务参数;
当所述特征参数位于预设的第二阈值区间时,确定所述样本用户具有所述第二业务参数;
其中,所述logistic回归分析模型是采用大量样本用户的特征数据进行logistic回归分析并反复迭代训练得到。
本申请的另一个目的是提供一种业务参数获取装置,所述装置包括:
获取单元,用于获取待预测业务参数的样本用户的特征数据;
处理单元,用于确定满足预置规则的样本用户为目标样本用户;
利用大量所述目标样本用户的特征数据确定所述logistic回归分析模型;
将所述特征数据输入到logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数;
当所述特征参数位于预设的第一阈值区间时,确定所述样本用户具有所述第一业务参数;
当所述特征参数位于预设的第二阈值区间时,确定所述样本用户具有所述第二业务参数,其中,所述logistic回归分析模型是采用大量样本用户的特征数据进行logistic回归分析并反复迭代训练得到。
本申请的再一个目的是提供一种业务参数获取设备,所述设备的结构中包括处理器和存储器,所述存储器用于存储支持数据处理的设备执行上述方法的程序,所述处理器被配置为用于执行所述存储器中存储的程序。所述数据库处理设备还可以包括通信接口,用于数据库处理设备与其他设备或通信网络通信。
本申请实施例提供了一种计算机存储介质,用于储存为上述业务参数获取装置所用的计算机软件指令,其包含用于执行上述方面为业务参数获取装置所设计的程序。
图1是本申请实施例业务参数获取方法的一种实施例的流程图;
图2是本申请实施例业务参数获取方法的另一种实施例的流程图;
图3是本申请实施例业务参数获取装置的一种实施例的结构图;
图4是本申请实施例业务参数获取装置的另一种实施例的结构图。
为了使本技术领域的人员更好地理解本申请方案,下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本申请一部分的实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都应当属于本申请保护的范围。
本申请的说明书和权利要求书及上述附图中的术语“第一”、“第二”、“第三”“第四”等是用于区别类似的对象,而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换,以便这里描述的实施例能够以除了在这里图示或描述的内容以外的顺序实施。此外,术语“包括”和“具有”以及他们的任何变形,意图在于覆盖不排他的包含,例如,包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元,而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。
在描述本申请实施例之前,先对本申请实施例中涉及到的名词做初步的介绍:
logistic回归分析模型是基于有监督训练的机器学习模型。
有监督学习:一种训练方法,有训练样本和训练标签。
在大学中经常会提到国家助学贷款,用于帮助家庭经济条件不好的学生完成学习,跟通常的贷款相类似的也是需要对在校生的违约进行预判,通过多种方式管控风险,例如延迟颁发毕业证或学位证等等,这些措施是在贷款之后对贷款人的风险管控措施,在进行贷款之前对于在校生的贷款违约的预测并没有做到很全面很准确。需要说明的是,本申请实施例的方案不限于社交应用,所有可以公开的用户特征数据都可以用作本申请实施例。
随着科技的发展,越来越多的社交应用走入我们生活,在用户授权的情况下,很多社交应用都可以将用户的所在位置和设备信息公布在社交圈,例如在朋友圈显示当前所在位置、在微博信息中标注发送微博设备的品牌型号,这些信息都可以体现出用户的特征数据,可以通过这些特征数据进行一些预
判。
本申请通过对用户的特征数据确定对应的业务参数,实际上这些业务参数可以反映用户在未来一段时间内的诚信情况,即是否会出现违约情况,本申请能反映用户是否能违约的业务参数可以是违约的概率,即在0到1之间,如果业务参数得到的违约概率更趋向于0则表明违约的可能性较小,例如违约概率为0.1,相反,若违约概率更趋向于1则表明违约的可能性较大,例如违约概率为0.9。本申请实施例中的违约预测和用户违约概率只是表达方式不同,实际上原理是相同的。
结合图1所示,针对以上传统方法及其缺点,本申请实施例提供了一种业务参数获取方法,所述方法包括:
S101、获取待预测业务参数的样本用户的特征数据。
S102、将所述特征数据输入到logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数,其中所述logistic回归分析模型是采用大量样本用户的特征数据进行logistic回归分析并反复迭代训练得到。
logistic回归分析模型中预先对大量的样本用户进行分析后可以得出样本用户的比较常用的特征参数,再对一个样本用户的业务参数继续获取时可以用logistic回归分析模型中已存在每种特征参数所对应的数值进行确定,因为对应样本用户可以有多种特征参数,每一种特征参数对于样本用户所对应的数值也不相同,例如,样本用户分别具有A特征参数、B特征参数及C特征参数,对应的数值可以分别为0.2、0.5及0.3,在确定该样本用户的时候可以利用特征参数进行确定业务参数,这里的业务参数可以代表的样本用户的信用程度,特征参数在0、1之间,如果特征参数趋向于1则表明违约的可能性较大,即信用度较低,反之,当特征参数趋向于0则表明违约的可能性较小,即信用度很高,通常可以选择中间值进行划分,例如将0到0.5之间作为第一阈值区间,0.5到1之间确定为第二阈值区间,当样本用户的特征参数处于第一阈值区间内则可以确定样本用户具有第一业务参数,当样本用户的特征参数位于第二阈值区间内则可以确定样本用户具有第二业务参数,因为logistic回归分析模型是预先对大量的样本用户进行分析后确定的特征参数对应的数值,这样对一个待测试业务参数的用户进行业务参数获取的时候结
果比较准确,能够较为客观对样本用户的违约进行预估。
结合图2所示,本申请实施例提供了一种业务参数获取方法,所述方法包括:
S201、确定满足预置规则的样本用户为目标样本用户。
预置规则至少包括:所述样本用户所处位置位于目标位置、与所述目标样本用户的关联程度达到预设关联阈值的用户、所述样本用户的身份信息符合预置条件,例如在进行在校生的违约预测时,可以利用在校生的所处位置和全国各大高校的地理位置进行匹配,对于在校生的所处位置可以使用设备的定位功能,对于在校生的所处位置应该在用户授权下获得,还可以进一步地利用年龄和/或网龄数据去除一部分不符合年龄的人群,因为在校生接触新事物比较多,对于上网时间会更多,通过对其社交媒体的账户等级也可以判断,对于确定为样本用户的在校生可以根据其关联的朋友圈进行衍生扩展出更多符合在校生条件的样本用户,这样在确定在校生的样本时可以有大量的样本供使用,提高logistic回归分析模型的准确性。
S202、利用大量所述目标样本用户的特征数据确定所述logistic回归分析模型。
对于特征参数可以包括对样本用户位置迁移频率、联系方式更新频率、社交应用信息的推送频率等进行统计分析,这些可以通过统计得到,再通过不断的重复迭代运算确定准确的特征参数以及这些参数对应的数值,即权重值,例如对一个人的位置迁移频率进行统计,出现的位置很多且不固定,可以认为该用户的工作或学习状态不稳定,向其分配业务时候,后期进展可能不会顺利,这样的特征参数再分配权重时可以提高该特征参数的权重值,体现出重要性。例如,对该用户进行贷款时,由于工作或学习不稳定,会产生不能按期还款的情况,这样的用户违约风险会提高,那么在进行贷款时进行更多的审查。
S203、获取待预测业务参数的样本用户的特征数据。
S204、将所述特征数据输入到logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数,其中,所述logistic回归分析模型是采用大量样本用户的特征数据进行logistic回归分析并反复迭代训练得到。
S205、当所述特征参数位于预设的第一阈值区间时,确定所述样本用户具有所述第一业务参数,当所述特征参数位于预设的第二阈值区间时,确定所述样本用户具有所述第二业务参数。
logistic回归分析模型根据特征数据输出的特征参数可以是一个概率值,特征参数的范围在0、1之间,将业务参数划分为两种类型包括第一业务参数和第二业务参数,第一业务参数还可以设定为诚信用户,第二业务参数可以设定为违约用户,当进行信用预测时候,业务参数可以对应用户的违约可能性,这样对应下来可以为诚信用户和违约用户,例如特征参数在0到0.5之间,此时样本用户具有诚信用户的特征更多,也可以说该样本用户违约的可能性较小,当特征参数在0.5到1之间时候,此时该样本用户具有违约用户的特征更多,可以说该样本用户违约的可能性较高,设置阈值区间时候可灵活选择,当需要判断诚信用户更严格,则可以将中间值的取值更靠近0,例如,第一阈值区间可以设定为0到0.2之间,而第二阈值区间对应设定在0.2到1之间,对应地,对诚信用户的条件宽松,则可以将中间值的取值更靠近1,例如,0.7,第一阈值区间可以设定为0到0.7,第二阈值区间可以设定为0.7到1,总之,通过特征参数的值可以确定样本用户的业务参数,可以对样本用户的违约情况进行预判。
本申请实施例中建立logistic回归分析模型的方法的一实施例包括
对所述目标样本用户的特征数据进行衍生并提取具有趋势性的第一参数;
对所述第一参数进行降维得到具有解释性的第二参数;
对所述第二参数依次进行聚类分析、判别分析以及去重以得到第三参数,其中,通过所述聚类分析从第二参数中选取N1个参数,通过所述判别分析从第二参数中选取N2个参数,将选取的N1个参数和N2个参数合并去重后得到第三参数;
对所述第三参数进行logistic回归分析以得到第四参数;
对所述第四参数进行重复迭代运算以得到模型参数,确定所述样本用户具有所述第二业务参数。
具体地说:根据Logistic函数的定义
logit(p)=α+β·X=α+β1x1+β2x2+...+βnxn
p事件发生的概率,β=(β1,β2,...,βn)为参数方程的估计值,X=(x1,x2,...,xn)T为logistic回归分析模型变量。
θ表示模型估计的参数,即:α,β1,β2,...,βn
因为y为二值分类,0或1,根据p1,p0这两个概率得出诚信用户和违约用户的概率分布情况。
p(y|x,θ)=(1-hθ(x))y·hθ(x)1-y
根据最大似然估计原理
通过对log(L(θ))求导,求出极值,得出θ的迭代函数,就是logistic回归分析模型估计参数,这里说的模型变量实际对应估计参数可以作为每个特征参数的权重值,在对一个用户进行预测时候,将该用户的特征数据进行分类得到多个特征参数,对多个特征参数配置权重值进行计算可以得到该用户的业务参数,即预估的违约概率,根据违约概率的数值可以对该用户的违约进行预估,以便决定是否对其执行相关业务,例如发放贷款等。
需要说明的是,logistic回归分析模型变量的选取的前提是衍生变量,通常作为分析的对象可以是用户或者帐户,所获得的数据可以有用户基本属性数据、社交属性数据、交易属性数据、稳定安全属性变量等等,可以根据这些数据进行衍生得到新的变量供使用,创建衍生变量的过程本领域普通技术人员应当了解,这里不进行赘述。
本申请实施例公开了一种业务参数获取方法,首先获取待预测业务参数
的样本用户的特征数据,将所述特征数据输入到logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数,其中,所述logistic回归分析模型是采用大量样本用户的特征数据进行logistic回归分析并反复迭代训练得到,因为logistic回归分析模型预先对大量的样本用户进行分析后确定的特征参数对应的数值,这样对一个待测试业务参数的用户进行业务参数获取时候结果比较准确,能够较为客观对样本用户的违约进行预估。
结合图3所示,前文中介绍了一种业务参数获取方法,对应地,本申请实施例中还提供一种业务参数获取装置,所述装置包括:
获取单元301,用于获取待预测业务参数的样本用户的特征数据;
分析单元302,用于利用logistic回归分析模型对所述特征数据进行归类分析,得到所述特征数据的多个特征参数;
获取单元301,用于获取待预测业务参数的样本用户的特征数据;
处理单元302,用于将所述特征数据输入到logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数,其中所述logistic回归分析模型是采用大量样本用户的特征数据进行logistic回归分析并反复迭代训练得到。
可选地,所述处理单元302还用于:
确定满足预置规则的样本用户为目标样本用户;
用于利用大量所述目标样本用户的特征数据确定所述logistic回归分析模型。
可选地,所述业务参数包括第一业务参数和第二业务参数,所述处理单元302还用于:
当所述特征参数位于预设的第一阈值区间时,确定所述样本用户具有所述第一业务参数;
当所述特征参数位于预设的第二阈值区间时,确定所述样本用户具有所述第二业务参数。
可选地,所述处理单元302还用于:
对所述目标样本用户的特征数据进行衍生并提取具有趋势性的第一参数;
对所述第一参数进行降维得到具有解释性的第二参数;
对所述第二参数依次进行聚类分析、判别分析以及去重以得到第三参数;
对所述第三参数进行logistic回归分析以得到第四参数;
对所述第四参数进行重复迭代运算以得到模型参数,确定所述样本用户具有所述第二业务参数。
可选地,所述预置规则至少包括:所述样本用户所处位置位于目标位置、与所述目标样本用户的关联程度达到预设关联阈值的用户、所述样本用户的身份信息符合预置条件。
本申请实施例公开了一种业务参数获取装置,首先获取待预测业务参数的样本用户的特征数据,利用logistic回归分析模型对所述特征数据进行归类分析,得到所述特征数据的多个特征参数,确定所述多个特征参数中的每一个特征参数的数值,所述数值用于确定所述业务参数,其中所述logistic回归分析模型是采用大量样本用户的特征数据进行logistic回归分析并反复迭代训练得到,因为logistic回归分析模型预先对大量的样本用户进行分析后确定的特征参数对应的数值,这样对一个待测试业务参数的用户进行业务参数获取时候结果比较准确,能够较为客观对样本用户的违约进行预估。
结合图4所示,图4是本申请实施例提供的业务参数获取装置40的结构示意图。所述业务参数获取装置40包括处理器410、存储器450和输入/输出I/O设备430,存储器450可以包括只读存储器和随机存取存储器,并向处理器410提供操作指令和数据。存储器450的一部分还可以包括非易失性随机存取存储器(NVRAM)。
在一些实施方式中,存储器450存储了如下的元素,可执行模块或者数据结构,或者他们的子集,或者他们的扩展集:
在本申请实施例中,通过调用存储器450存储的操作指令(该操作指令可存储在操作系统中),
获取待预测业务参数的样本用户的特征数据。
将所述特征数据输入到logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数,其中,所述logistic回归分析模型是采用大量样本用户的特征数据进行logistic回归分析并反复迭代训练得到。
处理器410控制业务参数获取装置40的操作,处理器410还可以称为
CPU(Central Processing Unit,中央处理单元)。存储器450可以包括只读存储器和随机存取存储器,并向处理器410提供指令和数据。存储器450的一部分还可以包括非易失性随机存取存储器(NVRAM)。的应用中业务参数获取装置40的各个组件通过总线系统420耦合在一起,其中总线系统420除包括数据总线之外,还可以包括电源总线、控制总线和状态信号总线等。但是为了清楚说明起见,在图中将各种总线都标为总线系统420。
上述本申请实施例揭示的方法可以应用于处理器410中,或者由处理器410实现。处理器410可能是一种集成电路芯片,具有信号的处理能力。在实现过程中,上述方法的各步骤可以通过处理器410中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器410可以是通用处理器、数字信号处理器(DSP)、专用集成电路(ASIC)、现成可编程门阵列(FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本申请实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本申请实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成,或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器,闪存、只读存储器,可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器450,处理器410读取存储器450中的信息,结合其硬件完成上述方法的步骤。
所属领域的技术人员可以清楚地了解到,为描述的方便和简洁,上述描述的系统,装置和单元的具体工作过程,可以参考前述方法实施例中的对应过程,在此不再赘述。
在本申请所提供的几个实施例中,应该理解到,所揭露的系统,装置和方法,可以通过其它的方式实现。例如,以上所描述的装置实施例仅仅是示意性的,例如,所述单元的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式,例如多个单元或组件可以结合或者可以集成到另一个系统,或一些特征可以忽略,或不执行。另一点,所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口,装置或单元的间接耦合或通信连接,可以是电性,机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的,作
为单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外,在本申请各个实施例中的各功能单元可以集成在一个处理单元中,也可以是各个单元单独物理存在,也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现,也可以采用软件功能单元的形式实现。
本领域普通技术人员可以理解上述实施例的各种方法中的全部或部分步骤是可以通过程序来指令相关的硬件来完成,该程序可以存储于一计算机可读存储介质中,存储介质可以包括:只读存储器(ROM,Read Only Memory)、随机存取存储器(RAM,Random Access Memory)、磁盘或光盘等。
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步骤是可以通过程序来指令相关的硬件完成,所述的程序可以存储于一种计算机可读存储介质中,上述提到的存储介质可以是只读存储器,磁盘或光盘等。
以上对本申请所提供的一种业务参数获取方法及装置进行了详细介绍,对于本领域的一般技术人员,依据本申请实施例的思想,在具体实施方式及应用范围上均会有改变之处,综上所述,本说明书内容不应理解为对本申请的限制。
Claims (9)
- 一种业务参数获取方法,其特征在于,所述方法包括:确定满足预置规则的样本用户为目标样本用户;利用所述目标样本用户的特征数据确定logistic回归分析模型;获取待预测业务参数的样本用户的特征数据;将所述特征数据输入到所述logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数;当所述特征参数位于预设的第一阈值区间时,确定所述样本用户具有所述第一业务参数;以及当所述特征参数位于预设的第二阈值区间时,确定所述样本用户具有所述第二业务参数;其中,所述logistic回归分析模型是采用样本用户的特征数据进行logistic回归分析并反复迭代训练得到。
- 根据权利要求1所述的方法,其特征在于,所述利用所述目标样本用户的特征数据确定所述logistic回归分析模型具体包括:对所述目标样本用户的特征数据进行衍生并提取具有趋势性的第一参数;对所述第一参数进行降维得到具有解释性的第二参数;对所述第二参数依次进行聚类分析、判别分析以及去重以得到第三参数;对所述第三参数进行logistic回归分析以得到第四参数;以及对所述第四参数进行重复迭代运算以得到模型参数,所述模型参数用于确定所述特征数据对应的所述特征参数。
- 根据权利要求1所述的方法,其特征在于,所述预置规则至少包括:所述样本用户所处位置位于目标位置、与所述目标样本用户的关联程度达到预设关联阈值的用户、所述样本用户的身份信息符合预置条件。
- 根据权利要求1所述的方法,其特征在于,所述第一阈值区间位于0和0.5之间,所述第二阈值区间位于0.5和1之间。
- 一种业务参数获取装置,其特征在于,所述装置包括:获取单元,用于获取待预测业务参数的样本用户的特征数据;处理单元,用于确定满足预置规则的样本用户为目标样本用户;利用所述目标样本用户的特征数据确定所述logistic回归分析模型;将所述特征数据输入到logistic回归分析模型得到所述特征数据的特征参数,所述特征参数用于确定所述业务参数;当所述特征参数位于预设的第一阈值区间时,确定所述样本用户具有所述第一业务参数;以及当所述特征参数位于预设的第二阈值区间时,确定所述样本用户具有所述第二业务参数,其中,所述logistic回归分析模型是采用样本用户的特征数据进行logistic回归分析并反复迭代训练得到。
- 根据权利要求5所述的装置,其特征在于,所述处理单元还用于:对所述目标样本用户的特征数据进行衍生并提取具有趋势性的第一参数;对所述第一参数进行降维得到具有解释性的第二参数;对所述第二参数依次进行聚类分析、判别分析以及去重以得到第三参数;对所述第三参数进行logistic回归分析以得到第四参数;以及对所述第四参数进行重复迭代运算以得到模型参数,所述模型参数用于确定所述特征数据对应的所述特征参数。
- 根据权利要求5所述的装置,其特征在于,所述预置规则至少包括:所述样本用户所处位置位于目标位置、与所述目标样本用户的关联程度达到预设关联阈值的用户、所述样本用户的身份信息符合预置条件。
- 一种业务参数获取设备,其特征在于,包括:处理器和存储器,其中,所述存储器中存有计算机可读程序;所述处理器通过运行所述存储器中的程序,以用于完成上述权利要求1至4所述的方法。
- 一种非易失性存储介质,用于存储一个或多个计算机程序,其中,所述计算机程序包括具有一个或多个存储器的处理器可运行的指令,所述指令被计算机执行时,使得所述计算机执行以下操作:确定满足预置规则的样本用户为目标样本用户;利用所述目标样本用户的特征数据确定logistic回归分析模型;获取待预测业务参数的样本用户的特征数据;将所述特征数据输入到所述logistic回归分析模型得到所述特征数据的 特征参数,所述特征参数用于确定所述业务参数;当所述特征参数位于预设的第一阈值区间时,确定所述样本用户具有所述第一业务参数;以及当所述特征参数位于预设的第二阈值区间时,确定所述样本用户具有所述第二业务参数;其中,所述logistic回归分析模型是采用样本用户的特征数据进行logistic回归分析并反复迭代训练得到。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610078384.8 | 2016-02-03 | ||
CN201610078384.8A CN107040397B (zh) | 2016-02-03 | 2016-02-03 | 一种业务参数获取方法及装置 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2017133615A1 true WO2017133615A1 (zh) | 2017-08-10 |
Family
ID=59500608
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2017/072593 WO2017133615A1 (zh) | 2016-02-03 | 2017-01-25 | 一种业务参数获取方法及装置 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN107040397B (zh) |
WO (1) | WO2017133615A1 (zh) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109005061A (zh) * | 2018-08-03 | 2018-12-14 | 深圳市科陆电子科技股份有限公司 | 参数管理方法、装置及存储介质 |
CN109685238A (zh) * | 2017-10-19 | 2019-04-26 | 腾讯科技(深圳)有限公司 | 资源交换方法和装置、存储介质及电子装置 |
CN110163713A (zh) * | 2019-01-28 | 2019-08-23 | 腾讯科技(深圳)有限公司 | 一种业务数据处理方法、装置以及相关设备 |
CN111274164A (zh) * | 2020-01-21 | 2020-06-12 | 苏州浪潮智能科技有限公司 | 一种lba分配方法、装置、设备及可读存储介质 |
CN111435452A (zh) * | 2019-01-11 | 2020-07-21 | 百度在线网络技术(北京)有限公司 | 模型训练方法、装置、设备和介质 |
CN111639117A (zh) * | 2020-05-26 | 2020-09-08 | 李绍兵 | 基于数据加工的业务处理方法及装置 |
CN112148765A (zh) * | 2019-06-28 | 2020-12-29 | 北京百度网讯科技有限公司 | 业务数据的处理方法、装置及存储介质 |
CN112445410A (zh) * | 2020-12-07 | 2021-03-05 | 北京小米移动软件有限公司 | 触控事件识别方法、装置及计算机可读存储介质 |
CN113051445A (zh) * | 2019-12-27 | 2021-06-29 | 北京国双科技有限公司 | 工业生产数据处理方法、装置、计算机设备和存储介质 |
CN113409084A (zh) * | 2017-10-19 | 2021-09-17 | 创新先进技术有限公司 | 模型训练方法、基于模型的用户行为预测方法及装置 |
CN113516333A (zh) * | 2021-03-10 | 2021-10-19 | 福建省农村信用社联合社 | 一种基于精准化业务模型的性能测试方法和系统 |
CN114862323A (zh) * | 2022-05-28 | 2022-08-05 | 平安银行股份有限公司 | 库存储备的分析方法、装置、设备及存储介质 |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107943571B (zh) * | 2017-11-14 | 2020-03-10 | Oppo广东移动通信有限公司 | 后台应用管控方法、装置、存储介质及电子设备 |
CN109871514B (zh) * | 2017-12-05 | 2022-11-04 | 财付通支付科技有限公司 | 一种数据处理方法、装置及存储介质 |
CN110957044A (zh) * | 2019-09-20 | 2020-04-03 | 上海派拉软件股份有限公司 | 基于改进的逻辑回归模型的健康管理方法 |
CN113537666B (zh) * | 2020-04-16 | 2024-05-03 | 马上消费金融股份有限公司 | 评测模型训练方法、评测和业务审核方法、装置及设备 |
CN112200272B (zh) * | 2020-12-07 | 2021-02-23 | 上海冰鉴信息科技有限公司 | 业务分类方法及装置 |
CN113887862A (zh) * | 2021-08-24 | 2022-01-04 | 国网天津市电力公司营销服务中心 | 一种能源计量业务数据分析方法和系统 |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101493913A (zh) * | 2008-01-23 | 2009-07-29 | 阿里巴巴集团控股有限公司 | 一种评估网上用户信用的方法及系统 |
US20120022945A1 (en) * | 2010-07-22 | 2012-01-26 | Visa International Service Association | Systems and Methods to Identify Payment Accounts Having Business Spending Activities |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI464700B (zh) * | 2011-10-31 | 2014-12-11 | Univ Ming Chuan | 信用違約預測方法與裝置 |
CN103970974A (zh) * | 2013-02-01 | 2014-08-06 | 无锡南理工科技发展有限公司 | 一种基于缺陷类别的安全风险评估方法 |
-
2016
- 2016-02-03 CN CN201610078384.8A patent/CN107040397B/zh active Active
-
2017
- 2017-01-25 WO PCT/CN2017/072593 patent/WO2017133615A1/zh active Application Filing
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101493913A (zh) * | 2008-01-23 | 2009-07-29 | 阿里巴巴集团控股有限公司 | 一种评估网上用户信用的方法及系统 |
US20120022945A1 (en) * | 2010-07-22 | 2012-01-26 | Visa International Service Association | Systems and Methods to Identify Payment Accounts Having Business Spending Activities |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109685238A (zh) * | 2017-10-19 | 2019-04-26 | 腾讯科技(深圳)有限公司 | 资源交换方法和装置、存储介质及电子装置 |
CN113409084A (zh) * | 2017-10-19 | 2021-09-17 | 创新先进技术有限公司 | 模型训练方法、基于模型的用户行为预测方法及装置 |
CN109005061A (zh) * | 2018-08-03 | 2018-12-14 | 深圳市科陆电子科技股份有限公司 | 参数管理方法、装置及存储介质 |
CN111435452A (zh) * | 2019-01-11 | 2020-07-21 | 百度在线网络技术(北京)有限公司 | 模型训练方法、装置、设备和介质 |
CN111435452B (zh) * | 2019-01-11 | 2023-11-03 | 百度在线网络技术(北京)有限公司 | 模型训练方法、装置、设备和介质 |
CN110163713A (zh) * | 2019-01-28 | 2019-08-23 | 腾讯科技(深圳)有限公司 | 一种业务数据处理方法、装置以及相关设备 |
CN112148765B (zh) * | 2019-06-28 | 2024-04-09 | 北京百度网讯科技有限公司 | 业务数据的处理方法、装置及存储介质 |
CN112148765A (zh) * | 2019-06-28 | 2020-12-29 | 北京百度网讯科技有限公司 | 业务数据的处理方法、装置及存储介质 |
CN113051445A (zh) * | 2019-12-27 | 2021-06-29 | 北京国双科技有限公司 | 工业生产数据处理方法、装置、计算机设备和存储介质 |
CN111274164B (zh) * | 2020-01-21 | 2022-07-08 | 苏州浪潮智能科技有限公司 | 一种lba分配方法、装置、设备及可读存储介质 |
CN111274164A (zh) * | 2020-01-21 | 2020-06-12 | 苏州浪潮智能科技有限公司 | 一种lba分配方法、装置、设备及可读存储介质 |
CN111639117A (zh) * | 2020-05-26 | 2020-09-08 | 李绍兵 | 基于数据加工的业务处理方法及装置 |
CN111639117B (zh) * | 2020-05-26 | 2023-12-01 | 四川三江数智科技有限公司 | 基于数据加工的业务处理方法及装置 |
CN112445410A (zh) * | 2020-12-07 | 2021-03-05 | 北京小米移动软件有限公司 | 触控事件识别方法、装置及计算机可读存储介质 |
CN112445410B (zh) * | 2020-12-07 | 2023-04-18 | 北京小米移动软件有限公司 | 触控事件识别方法、装置及计算机可读存储介质 |
CN113516333A (zh) * | 2021-03-10 | 2021-10-19 | 福建省农村信用社联合社 | 一种基于精准化业务模型的性能测试方法和系统 |
CN113516333B (zh) * | 2021-03-10 | 2023-11-14 | 福建省农村信用社联合社 | 一种基于精准化业务模型的性能测试方法和系统 |
CN114862323A (zh) * | 2022-05-28 | 2022-08-05 | 平安银行股份有限公司 | 库存储备的分析方法、装置、设备及存储介质 |
Also Published As
Publication number | Publication date |
---|---|
CN107040397B (zh) | 2020-12-11 |
CN107040397A (zh) | 2017-08-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2017133615A1 (zh) | 一种业务参数获取方法及装置 | |
US20180253657A1 (en) | Real-time credit risk management system | |
US10504120B2 (en) | Determining a temporary transaction limit | |
CN109543925B (zh) | 基于机器学习的风险预测方法、装置、计算机设备和存储介质 | |
WO2018040068A1 (zh) | 基于知识图谱的语意分析系统及方法 | |
WO2020176977A1 (en) | Multi-page online application origination (oao) service for fraud prevention systems | |
US20180225450A1 (en) | Counter-fraud operation management | |
US20210233080A1 (en) | Utilizing a time-dependent graph convolutional neural network for fraudulent transaction identification | |
AU2017101862A4 (en) | Collaborative filtering method, apparatus, server and storage medium in combination with time factor | |
US10504028B1 (en) | Techniques to use machine learning for risk management | |
CN108520041B (zh) | 文本的行业分类方法、系统、计算机设备和存储介质 | |
US20130246290A1 (en) | Machine-Assisted Legal Assessments | |
WO2020211357A1 (zh) | 数据的关联分析方法、装置、计算机设备及存储介质 | |
JP2017535857A (ja) | 変換されたデータを用いた学習 | |
CN110705719A (zh) | 执行自动机器学习的方法和装置 | |
US11323564B2 (en) | Case management virtual assistant to enable predictive outputs | |
US20160055496A1 (en) | Churn prediction based on existing event data | |
JP2017527013A (ja) | サービスとしての適応特徴化 | |
US20210357699A1 (en) | Data quality assessment for data analytics | |
CN112348321A (zh) | 风险用户的识别方法、装置及电子设备 | |
US12099631B2 (en) | Rule-based anonymization of datasets | |
Boz et al. | Reassessment and monitoring of loan applications with machine learning | |
CN114638695A (zh) | 信用评估方法、装置、设备及介质 | |
CN111191677B (zh) | 用户特征数据生成方法、装置及电子设备 | |
US20220076157A1 (en) | Data analysis system using artificial intelligence |
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: 17746935 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: 17746935 Country of ref document: EP Kind code of ref document: A1 |