CN114612011A - Risk prevention and control decision method and device - Google Patents
Risk prevention and control decision method and device Download PDFInfo
- Publication number
- CN114612011A CN114612011A CN202210418916.3A CN202210418916A CN114612011A CN 114612011 A CN114612011 A CN 114612011A CN 202210418916 A CN202210418916 A CN 202210418916A CN 114612011 A CN114612011 A CN 114612011A
- Authority
- CN
- China
- Prior art keywords
- risk
- event data
- risk event
- sample
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 63
- 230000002265 prevention Effects 0.000 title claims abstract description 56
- 238000012549 training Methods 0.000 claims abstract description 31
- 238000004088 simulation Methods 0.000 claims abstract description 28
- 238000011217 control strategy Methods 0.000 claims abstract description 26
- 230000002159 abnormal effect Effects 0.000 claims abstract description 22
- 238000010801 machine learning Methods 0.000 claims abstract description 18
- 238000012545 processing Methods 0.000 claims description 81
- 230000006399 behavior Effects 0.000 claims description 58
- 230000006855 networking Effects 0.000 claims description 55
- 238000004364 calculation method Methods 0.000 claims description 28
- 238000007781 pre-processing Methods 0.000 claims description 27
- 238000004590 computer program Methods 0.000 claims description 11
- 238000011156 evaluation Methods 0.000 claims description 9
- 238000012216 screening Methods 0.000 claims description 7
- 238000012821 model calculation Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 description 11
- 230000001133 acceleration Effects 0.000 description 6
- 238000013473 artificial intelligence Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 238000013475 authorization Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Strategic Management (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Development Economics (AREA)
- Medical Informatics (AREA)
- Educational Administration (AREA)
- Evolutionary Biology (AREA)
- Game Theory and Decision Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The disclosure provides a risk prevention and control decision method and a risk prevention and control decision device. Wherein, the method comprises the following steps: determining risk event data to be prevented and controlled; inputting the risk event data to be prevented and controlled into a preset risk decision model to obtain a target wind control strategy output by the risk decision model; the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data; the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data. According to the risk prevention and control decision method, the risk decision model obtained through training of the abnormal risk event data and the historical risk event data is used for quickly sensing and identifying the risk event data to be prevented and controlled, and an effective wind control strategy is generated, so that the accuracy of wind control decision is improved.
Description
Technical Field
The disclosure relates to the technical field of computer security, in particular to a risk prevention and control decision method and device. In addition, an electronic device and a processor-readable storage medium are also related.
Background
With the rapid development of artificial intelligence technology, the research of intelligent decision-making systems is more and more intensive. At present, in the field of financial and automated decision making, an intelligent decision making system has realized technologies such as risk perception, risk identification, intelligent evolution and the like to a certain extent, and has been widely applied in the actual situation. High-quality data such as user basic information, user behavior information, user authorization information, external access information and the like are required, the data sources are summarized, the decision on credit risk, fraud risk and system risk is achieved through artificial intelligence model calculation, and the decision is provided for corresponding professional business personnel to make decisions. However, the training data of the existing artificial intelligence model is relatively single, and is usually directly acquired historical data, so that the decision accuracy of the model is poor. Therefore, how to accurately make a risk prevention and control decision becomes an urgent problem to be solved.
Disclosure of Invention
Therefore, the risk prevention and control decision method and device are provided to solve the defect that the accuracy of the risk prevention and control decision is poor due to the fact that an artificial intelligence model for the risk prevention and control decision in the prior art is high in limitation.
The present disclosure provides a risk prevention and control decision method, comprising:
determining risk event data to be prevented and controlled;
inputting the risk event data to be prevented and controlled into a preset risk decision model to obtain a target wind control strategy output by the risk decision model;
the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data;
the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data.
Further, the risk prevention and control decision method further includes: predetermining sample risk event data;
the predetermined sample risk event data specifically includes:
acquiring initial sample risk event data;
determining a target networking service terminal in the networking service terminals based on the user use state, the equipment operation state, the equipment performance parameters and the equipment load information of the networking service terminals; sending the initial sample risk event data to at least one corresponding target networking service terminal for preprocessing; and/or acquiring sample data obtained by locally acquiring data and preprocessing at least one target networking service terminal to obtain preprocessed sample data;
and carrying out sample mixing on the preprocessed sample data, and selecting the sample risk event data meeting preset conditions in a sample fine adjustment and screening processing mode.
Further, the abnormal risk event data obtained through simulation based on the historical risk event data specifically includes:
marking features in the historical risk event data;
and analyzing the marked features by using a simulation model based on the event features corresponding to the typical samples in the historical risk event data to generate simulated typical samples corresponding to the typical samples, and taking the simulated typical samples as abnormal risk event data.
Further, before the risk prevention and control decision method inputs the risk event data to be prevented and controlled into a preset risk decision model, the method further includes:
inputting the risk event data to be prevented and controlled into a preset scheduling module to determine a processing type corresponding to the risk event data to be prevented and controlled, and distributing corresponding computing resources based on the processing type; processing a corresponding risk decision model calculation task by utilizing the calculation force resource so as to realize the processing of the risk event data to be prevented and controlled;
the scheduling module is used for matching corresponding computing resources according to the processing type required by the risk event data to be prevented and controlled; the processing type includes at least one of real-time processing, asynchronous processing, and offline processing.
Further, the risk prevention and control decision method further includes:
determining operation behavior data corresponding to the risk event data to be prevented and controlled;
analyzing the operation behavior data by using a system decision model in a risk prevention and control system to obtain a behavior evaluation result of the operation behavior data; the system decision model is a machine learning model obtained by performing model training by taking historical operation behavior data and processing results corresponding to the operation behavior data as training samples.
Further, the operation behavior data includes at least one of a previous operation behavior prediction data, a previous operation behavior data, and a subsequent operation behavior data.
The present disclosure also provides a risk prevention and control decision device, including:
the risk event data determining unit is used for determining risk event data to be prevented and controlled;
the wind control strategy determining unit is used for inputting the risk event data to be prevented and controlled into a preset risk decision model to obtain a target wind control strategy output by the risk decision model;
the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data;
the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data.
Further, the risk prevention and control decision device further includes: a sample preprocessing unit; the sample preprocessing unit is specifically configured to: acquiring initial sample risk event data; determining a target networking service terminal in the networking service terminals based on the user use state, the equipment operation state, the equipment performance parameters and the equipment load information of the networking service terminals; sending the initial sample risk event data to at least one corresponding target networking service terminal for preprocessing; and/or acquiring sample data obtained by locally acquiring data and preprocessing at least one target networking service terminal to obtain preprocessed sample data; and carrying out sample mixing on the preprocessed sample data, and selecting the sample risk event data meeting preset conditions in a sample fine adjustment and screening processing mode.
Further, the abnormal risk event data obtained through simulation based on the historical risk event data specifically includes:
marking features in the historical risk event data;
and analyzing the marked features by using a simulation model based on the event features corresponding to the typical samples in the historical risk event data to generate simulated typical samples corresponding to the typical samples, and taking the simulated typical samples as abnormal risk event data.
Further, the risk prevention and control decision device further includes: the scheduling processing unit is used for inputting the risk event data to be prevented and controlled into a preset scheduling module so as to determine a processing type corresponding to the risk event data to be prevented and controlled, and distributing corresponding computing resources based on the processing type; processing a corresponding risk decision model calculation task by utilizing the calculation force resource so as to realize the processing of the risk event data to be prevented and controlled;
the scheduling module is used for matching corresponding computing resources according to the processing type required by the risk event data to be prevented and controlled; the processing type includes at least one of real-time processing, asynchronous processing, and offline processing.
Further, the risk prevention and control decision device further includes:
the operation behavior data determining unit is used for determining operation behavior data corresponding to the risk event data to be prevented and controlled;
the operation behavior evaluation unit is used for analyzing the operation behavior data by utilizing a system decision model in the risk prevention and control system to obtain a behavior evaluation result of the operation behavior data; the system decision model is a machine learning model obtained by performing model training by taking historical operation behavior data and a processing result corresponding to the operation behavior data as training samples.
Further, the operation behavior data includes at least one of a previous operation behavior prediction data, a previous operation behavior data, and a subsequent operation behavior data.
The present disclosure also provides an electronic device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the risk prevention and control decision method as described in any one of the above when executing the program.
The present disclosure also provides a processor-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the risk prevention and control decision method as described in any one of the above.
According to the risk prevention and control decision method, the risk event data to be prevented and controlled are determined and input into the preset risk decision model, so that the risk event data to be prevented and controlled are effectively sensed and identified, and the corresponding target wind control strategy is generated, and the accuracy of the wind control decision is improved.
Drawings
In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present disclosure, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a risk prevention and control decision method provided by an embodiment of the present disclosure;
FIG. 2 is a complete flow chart diagram of a risk prevention and control decision method provided by an embodiment of the present disclosure;
FIG. 3 is a schematic diagram of a sample pre-processing process provided by embodiments of the present disclosure;
FIG. 4 is a schematic diagram of a data sample processing process provided by an embodiment of the present disclosure;
FIG. 5 is a schematic diagram of a risk decision model training process provided by an embodiment of the present disclosure;
FIG. 6 is a schematic diagram of a risk prevention decision process provided by an embodiment of the present disclosure;
FIG. 7 is a schematic diagram of a target wind control strategy analysis process provided by an embodiment of the present disclosure;
FIG. 8 is a schematic diagram of a system decision process provided by an embodiment of the present disclosure;
fig. 9 is a schematic structural diagram of a risk prevention and control decision device provided in an embodiment of the present disclosure;
fig. 10 is a schematic physical structure diagram of an electronic device provided in an embodiment of the present disclosure.
Detailed Description
To make the objects, technical solutions and advantages of the embodiments of the present disclosure more clear, the technical solutions of the embodiments of the present disclosure will be described clearly and completely with reference to the drawings in the embodiments of the present disclosure, and it is obvious that the described embodiments are some, but not all embodiments of the present disclosure. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.
The embodiment of the risk prevention and control decision method based on the present disclosure is described in detail below. As shown in fig. 1, which is a schematic flow chart of a risk prevention and control decision method provided in the embodiment of the present disclosure, a specific implementation process includes the following steps:
step 101: and determining risk event data to be prevented and controlled.
In the embodiment of the present disclosure, before executing this step, sample risk event data needs to be predetermined, and model training is performed to obtain the risk decision model. Specifically, initial sample risk event data is obtained firstly, a target networking service terminal in the networking service terminals is determined based on the user use state, the equipment operation state, the equipment performance parameters and the equipment load information of the networking service terminals, and the initial sample risk event data is sent to at least one corresponding target networking service terminal for sample preprocessing; and/or acquiring sample data obtained by locally acquiring data and preprocessing at least one target networking service terminal to obtain preprocessed sample data; and further, carrying out sample mixing on the preprocessed sample data, and selecting the sample risk event data meeting preset conditions in a sample fine adjustment and screening processing mode. The networking service terminal can be a terminal with driving force, such as a networking device terminal mainly comprising an electric automobile, a vehicle-mounted mobile workstation and other internet-of-things terminals; the terminal can also be a non-driving force terminal, such as an internet of things terminal like a smart phone and wearable networking equipment.
As shown in fig. 2, before model training, initial sample risk event data needs to be obtained in advance through user data partial features, model data and corresponding external data sources, sample preprocessing is performed on the initial sample risk event data, and a final sample (i.e., the sample risk event data meeting preset conditions) having characteristics of representativeness, sufficiency, timeliness, exclusiveness and the like is extracted from the initial sample risk event data, so that effective training sample data support is provided for a training process of a risk decision model.
In a specific implementation process, a server side can partially release service resources, the service resources are not complexity calculation generally but belong to a mild calculation scene, and the service resources can be put on a networking service terminal (such as a terminal with driving force) for implementation; or some service resources of the application program itself on the networked service terminal, and the service resources of the part which should be on the server side are handed over to the networked service terminal (such as a terminal with driving force or a terminal without driving force). The business resources comprise initial sample risk event data which are derived from partial characteristics of user data, model data and corresponding external data sources.
And determining the implementation process of a target networking service terminal in the networking service terminals aiming at the user use state, the equipment operation state, the equipment performance parameters and the equipment load information based on the networking service terminals. As shown in fig. 3, in the embodiment of the present disclosure, it needs to determine whether the networking service terminal device hardware can have the condition for executing the service resource based on several predetermined dimensions, such as whether the device is in an executable list, whether a local corresponding chip of the networking service terminal meets the calculation requirement, and the like, where the determination process mainly aims at the networking service terminal device hardware, and the determination standard is provided by the server in real time; after the judgment of the hardware level of the networking service terminal equipment is finished, the time zone in which the networking service terminal is positioned can be judged, whether the current time zone is in the night range or not is judged, and when the time zone is in the night range, the current running state of the networking service terminal equipment is judged in the next step, specifically comprising the network quality, the load state of the networking service terminal equipment and the like, so as to judge whether the requirement is met or not; in addition, the method also comprises the steps of judging whether the network belongs to a mobile network or a fixed network, if the network is the fixed network, detecting the link quality, and if the current network quality meets the requirement, starting to detect the running state of the networking service terminal equipment and the like, such as whether the Central Processing Unit (CPU) and the memory utilization rate meet the requirement. The device load status includes CPU occupancy and Random Access Memory (RAM) occupancy. If the current equipment load state is below a standard threshold value provided by the server, entering battery state judgment; for the networking service terminal equipment supporting the battery, judging whether the battery state is normal or not, if no problem exists, judging whether the battery state is in a charging state, and if the equipment is in the charging state, detecting the motion state of the networking service terminal equipment; the motion state also needs history and real-time judgment, the history is detected by low frequency and is stored locally, whether the equipment is in the motion state can be judged by relative sensors such as a 6-axis accelerator/gyroscope, if the equipment is not in the motion state, whether the equipment is close to the horizontal position or other fixed positions is judged, and other fixed positions need to be compared with historical data, and whether the equipment is in a normal fixed position is judged. After a series of detection of the user use state, the equipment running state, the equipment performance parameters, the equipment load information and the like is completed, when the current equipment (the networking service terminal) is in a state of being static, standby, charged, having a small load and the computing performance conforming to a preset condition, the capability of executing the computing service resources can be determined, enough computing resources can be provided for sample preprocessing analysis, and the networking service terminal with the capability of executing the computing service resources is determined to be a target networking service terminal.
In an implementation process of sending the initial sample risk event data to at least one corresponding target networked service terminal for preprocessing to obtain preprocessed sample data, the target networked service terminal may be a terminal with driving force or a terminal without driving force, such as an electric vehicle, a wearable device, a smartphone, or the like. By reasonably utilizing the computing resources of the target networking service terminal in the Internet of things, the computing load of sample preprocessing can be reduced, the network transmission time and the data volume are reduced, effective data training samples are analyzed in a differentiated mode, and therefore the quality of the preprocessed samples is greatly improved.
It should be noted that, with the introduction of a large amount of streaming media data, the data magnitude is larger and larger, the amount of sample data for model processing is larger and larger, the logic type and vision type calculations occur at the later stage of model processing, typical samples become smaller and smaller, the space for improving the algorithm is lower and lower, and if a large amount of homogeneous samples are still relied on, only learning the traditional type samples has extremely low energy efficiency, so when sample data is collected, the typical samples are screened, and more simulated typical samples are obtained through simulation, thereby improving the processing accuracy of the calculation tasks of the model in various scenes. As described above, there are fewer typical samples to the end, and the typical sample data actually collected is very low in percentage. Therefore, in the embodiment of the present disclosure, the abnormal risk event data obtained based on the historical risk event data through simulation includes: marking features in the historical risk event data; and analyzing the marked features by using a simulation model based on the event features corresponding to the typical samples in the historical risk event data to generate simulated typical samples corresponding to the typical samples, and taking the simulated typical samples as abnormal risk event data. By labeling the characteristics in each historical risk event data in the sample risk event data and performing characteristic extraction on the collected typical samples, according to the characteristic information of the typical samples, calculating through an analog simulation model according to the upper and lower preset intervals of the characteristic dimension of the typical samples, analyzing the marked features, generating simulation typical samples in real time by taking the real world as a reference, providing the simulation typical samples for the machine learning model (namely, a risk decision model) to carry out training learning (as shown in FIG. 5), meanwhile, an evaluation system is established for the machine learning model, corresponding operational thinking logic is generated through scoring, universal samples and typical samples in the real environment are dynamically adjusted, characteristic dimension intervals and magnitude division of the typical samples in the simulation environment are achieved, and a risk decision model meeting preset conditions is continuously output through the evaluation system. By the method, a large number of simulation typical samples can be split according to part of the typical samples, unnecessary performance dependence on network and server performance is reduced, and different combinations are performed on all characteristics in the real sample data through the simulation model, so that the problem of sample unicity existing in the collected historical risk event data is solved, the types of training samples of the risk decision model are enriched, and the prediction accuracy of the risk decision model is improved.
After the sample preprocessing stage is carried out to obtain preprocessed sample data, sample mixing is carried out on the preprocessed sample data, and the sample risk event data meeting preset conditions are selected in a sample fine adjustment and screening processing mode. In a specific implementation process, sample mixing is firstly carried out on preprocessed sample data, and then final sample processing is completed through a data sample processing stage to obtain sample risk event data containing historical risk event data and abnormal risk event data obtained based on historical risk event data simulation. As shown in fig. 4, the sample is mixed with several conditions: in the first case, if the sample processing strategy of the target networking service terminal is changed, discarding the preprocessed samples of the corresponding batch of target networking service terminals, and recalculating by the server; in case two, the source processing is passed, and after sample mixing, the target networking service terminal preprocessing sample and the server terminal processing sample meet the requirements, and then the data processing sample stage shown in fig. 2 is uniformly entered; in the third case, the source processing is carried out, and the target networking service terminal preprocessing or the server side sample needs to be finely adjusted through sample mixing; and selecting a proper sample by methods of representativeness, sufficiency, timeliness, exclusivity and the like, matching effective all information as basic characteristics, and finishing data sample processing.
Step 102: inputting the risk event data to be prevented and controlled into a preset risk decision model to obtain a target wind control strategy output by the risk decision model; the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data; the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data.
As shown in fig. 6, in the embodiment of the present disclosure, before inputting the risk event data to be controlled into a preset risk decision model, the method further includes: firstly, inputting the risk event data to be prevented and controlled into a preset scheduling module (namely a scheduling system) to determine a processing type corresponding to the risk event data to be prevented and controlled, allocating corresponding computational resources based on the processing type, and processing corresponding risk decision model calculation tasks by using the computational resources to realize the processing of the risk event data to be prevented and controlled. The scheduling module is used for matching corresponding computing resources according to the processing type required by the risk event data to be prevented and controlled; the processing type includes at least one of real-time processing, asynchronous processing, and offline processing.
Specifically, it is first determined whether the processing type of the model belongs to real-time processing, asynchronous processing, or offline processing, and a corresponding determination calculation scenario is determined. The calculation scene is one of a neural network calculation scene and a non-neural network calculation scene. And when the calculation scene is a neural network calculation scene, performing calculation processing through a deep learning model. And matching the computing hardware environment required by the model through the computing scene and the processing type to determine corresponding computing power resources. Specifically, the intelligent type selection can be performed on the computing chip according to the actual service situation. As shown in fig. 6, according to the calculation type (processing type), the CPU for scalar calculation, the Graphic Processor (GPU) for vector calculation, the Application Specific Integrated Circuit (ASIC) for matrix calculation, and the fpga (field Programmable Gate array) for spatial calculation can be classified. In the specific implementation process, the service resources are pre-estimated in advance according to the calculation type, and the acceleration type which is adopted is determined by the intelligent AI according to the suitable service resource type, wherein the acceleration type comprises an acceleration module, an acceleration card, an acceleration server and an acceleration calculation cluster, so that the load balance of each operation cluster on the whole is effectively ensured. According to the service type and the future service growth trend, different chips are combined through a scheduling system, and the optimal balance point is determined, so that more and more complex calculation tasks can be efficiently completed under various conditions.
As shown in fig. 7, after the target wind control policy generated by the risk decision model, a policy scheme is generated according to information fed back by user experience and risk target configuration, and a corresponding report is generated by policy pre-execution to determine whether the policy is in accordance with expectations, and if so, an effective target wind control policy is output.
In addition, it should be noted that in the sample preprocessing, model training, and model decision processing processes, accidents may actually occur at a certain probability regardless of human or non-human, so the embodiment of the present disclosure further includes determining, by a decision system, operation behavior data corresponding to the risk event data to be prevented and controlled. Specifically, the operation behavior data is analyzed by using a system decision model in the risk prevention and control system to obtain a behavior evaluation result of the operation behavior data, so that the occurrence probability of the same type of accidents is effectively reduced through a machine learning model on the premise of not spending excessive computing resources. The system decision model is a machine learning model obtained by performing model training by taking historical operation behavior data and a processing result corresponding to the operation behavior data as training samples. The operational behavior data may include at least one of prior operational behavior prediction data, prior operational behavior data, and post operational behavior data. As shown in fig. 8, in the embodiment of the present disclosure, the system decision part mainly includes a core logic intelligence AI to ensure that the risk of system configuration is minimized. Model learning and prediction are carried out through intelligent AI, and with the gradual increase of operation data of positive and negative samples, effective solutions can be provided in advance for operation prediction and in-process perception prompt. The previous operation prediction data mainly come from operation behavior data of daily users, and an operation behavior sequence of each user is evaluated, and the operation behavior sequence adopts a real-time calculation scheme to provide effective prediction data and prompt the user for operation risks in time; the data of the system operation interval between the operation data source operation and the operation in the process comprises multi-level external data sources, is suitable for adopting asynchronous and offline calculation schemes, and simultaneously ensures that the offline calculated data is within a preset time threshold value, thereby ensuring multiple offline calculations which may be generated in a short time; the post-operation data is from the data of a pre-event timeline and the data of a historical post-event processing scheme, and the post-operation data does not relate to an external data source basically, is mainly data inside the system and adopts a real-time calculation scheme.
According to the risk prevention and control decision method, the risk event data to be prevented and controlled are determined and input into the preset risk decision model, so that the risk event data to be prevented and controlled are effectively sensed and identified, a corresponding target wind control strategy is generated, and accuracy of wind control decision is improved.
Corresponding to the risk prevention and control decision method, the disclosure also provides a risk prevention and control decision device. Since the embodiment of the device is similar to the embodiment of the method described above, the description is simple, and for the relevant points, reference may be made to the description of the embodiment of the method described above, and the embodiment of the risk prevention and control decision device described below is only illustrative. Please refer to fig. 9, which is a schematic structural diagram of a risk prevention and control decision device according to an embodiment of the present disclosure.
The risk prevention and control decision device specifically comprises the following parts:
a risk event data determining unit 901, configured to determine risk event data to be prevented and controlled;
a wind control policy determining unit 902, configured to input the risk event data to be prevented and controlled into a preset risk decision model, so as to obtain a target wind control policy output by the risk decision model; the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data;
the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data.
Further, the risk prevention and control decision device further includes: a sample preprocessing unit; the sample preprocessing unit is specifically configured to: acquiring initial sample risk event data; determining a target networking service terminal in the networking service terminals based on the user use state, the equipment operation state, the equipment performance parameters and the equipment load information of the networking service terminals; sending the initial sample risk event data to at least one corresponding target networking service terminal for preprocessing; and/or acquiring sample data obtained by locally acquiring data and preprocessing at least one target networking service terminal to obtain preprocessed sample data; and carrying out sample mixing on the preprocessed sample data, and selecting the sample risk event data meeting preset conditions in a sample fine adjustment and screening processing mode.
Further, the abnormal risk event data obtained through simulation based on the historical risk event data specifically includes:
marking features in the historical risk event data;
and analyzing the marked features by using a simulation model based on the event features corresponding to the typical samples in the historical risk event data to generate simulated typical samples corresponding to the typical samples, and taking the simulated typical samples as abnormal risk event data.
Further, the risk prevention and control decision device further includes: the scheduling processing unit is used for inputting the risk event data to be prevented and controlled into a preset scheduling module so as to determine a processing type corresponding to the risk event data to be prevented and controlled, and distributing corresponding computational resources based on the processing type; processing a corresponding risk decision model calculation task by utilizing the calculation force resource so as to realize the processing of the risk event data to be prevented and controlled;
the scheduling module is used for matching corresponding computing resources according to the processing type required by the risk event data to be prevented and controlled; the processing type includes at least one of real-time processing, asynchronous processing, and offline processing.
Further, the risk prevention and control decision device further includes:
the operation behavior data determining unit is used for determining operation behavior data corresponding to the risk event data to be prevented and controlled;
the operation behavior evaluation unit is used for analyzing the operation behavior data by utilizing a system decision model in the risk prevention and control system to obtain a behavior evaluation result of the operation behavior data; the system decision model is a machine learning model obtained by performing model training by taking historical operation behavior data and a processing result corresponding to the operation behavior data as training samples.
Further, the operation behavior data includes at least one of a previous operation behavior prediction data, a previous operation behavior data, and a subsequent operation behavior data.
According to the risk prevention and control decision-making device, the risk event data to be prevented and controlled are determined and input into the preset risk decision-making model, so that the risk event data to be prevented and controlled are effectively sensed and identified, and a corresponding target wind control strategy is generated, and the accuracy of wind control decision-making is improved.
Corresponding to the risk prevention and control decision method, the disclosure also provides an electronic device. Since the embodiment of the electronic device is similar to the above method embodiment, the description is simple, and please refer to the description of the above method embodiment, and the electronic device described below is only schematic. Fig. 10 is a schematic physical structure diagram of an electronic device according to an embodiment of the disclosure. The electronic device may include: a processor (processor)1001, a memory (memory)1002, and a communication bus 1003, wherein the processor 1001 and the memory 1002 communicate with each other through the communication bus 1003, and communicate with the outside through the communication interface 1004. The processor 1001 may invoke logic instructions in the memory 1002 to perform a risk prevention decision method comprising: determining risk event data to be prevented and controlled; inputting the risk event data to be prevented and controlled into a preset risk decision model to obtain a target wind control strategy output by the risk decision model; the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data; the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data.
In addition, the logic instructions in the memory 1002 may be implemented in software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present disclosure may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a computer, or a network device) to execute all or part of the steps of the method according to the embodiments of the present disclosure. And the aforementioned storage medium includes: various media capable of storing program codes, such as a Memory chip, a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
In another aspect, the embodiments of the present disclosure also provide a computer program product, which includes a computer program stored on a processor-readable storage medium, the computer program including program instructions, when the program instructions are executed by a computer, the computer being capable of executing the risk prevention and control decision method provided by the above-mentioned method embodiments. The method comprises the following steps: determining risk event data to be prevented and controlled; inputting the risk event data to be prevented and controlled into a preset risk decision model to obtain a target wind control strategy output by the risk decision model; the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data; the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data.
In another aspect, the disclosed embodiments also provide a processor-readable storage medium, on which a computer program is stored, where the computer program is implemented to perform the risk prevention and control decision method provided by the above embodiments when executed by a processor. The method comprises the following steps: determining risk event data to be prevented and controlled; inputting the risk event data to be prevented and controlled into a preset risk decision model to obtain a target wind control strategy output by the risk decision model; the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data; the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data.
The processor-readable storage medium can be any available medium or data storage device that can be accessed by a processor, including, but not limited to, magnetic memory (e.g., floppy disks, hard disks, magnetic tape, magneto-optical disks (MOs), etc.), optical memory (e.g., CDs, DVDs, BDs, HVDs, etc.), and semiconductor memory (e.g., ROMs, EPROMs, EEPROMs, non-volatile memory (NAND FLASH), Solid State Disks (SSDs)), etc.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a computer, or a network device) to execute the method according to the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solutions of the present disclosure, not to limit them; although the present disclosure has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present disclosure.
Claims (10)
1. A risk prevention and control decision method is characterized by comprising the following steps:
determining risk event data to be prevented and controlled;
inputting the risk event data to be prevented and controlled into a preset risk decision model to obtain a target wind control strategy output by the risk decision model; the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data; the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data.
2. The risk prevention and control decision method of claim 1, further comprising: predetermining sample risk event data;
the predetermined sample risk event data specifically includes:
acquiring initial sample risk event data;
determining a target networking service terminal in the networking service terminals based on the user use state, the equipment operation state, the equipment performance parameters and the equipment load information of the networking service terminals; sending the initial sample risk event data to at least one corresponding target networking service terminal for preprocessing; and/or acquiring sample data obtained by locally acquiring data and preprocessing at least one target networking service terminal to obtain preprocessed sample data;
and carrying out sample mixing on the preprocessed sample data, and selecting the sample risk event data meeting preset conditions in a sample fine adjustment and screening processing mode.
3. The risk prevention and control decision method according to claim 1, wherein the abnormal risk event data obtained based on the historical risk event data simulation specifically includes:
marking features in the historical risk event data;
and analyzing the marked features by using a simulation model based on the event features corresponding to the typical samples in the historical risk event data to generate simulated typical samples corresponding to the typical samples, and taking the simulated typical samples as abnormal risk event data.
4. The risk prevention and control decision method according to claim 1, further comprising, before inputting the risk event data to be prevented and controlled into a preset risk decision model:
inputting the risk event data to be prevented and controlled into a preset scheduling module to determine a processing type corresponding to the risk event data to be prevented and controlled, and distributing corresponding computing resources based on the processing type; processing a corresponding risk decision model calculation task by utilizing the calculation force resource so as to realize the processing of the risk event data to be prevented and controlled;
the scheduling module is used for matching corresponding computing resources according to the processing type required by the risk event data to be prevented and controlled; the processing type includes at least one of real-time processing, asynchronous processing, and offline processing.
5. The risk prevention and control decision method of claim 1, further comprising:
determining operation behavior data corresponding to the risk event data to be prevented and controlled;
analyzing the operation behavior data by using a system decision model in a risk prevention and control system to obtain a behavior evaluation result of the operation behavior data; the system decision model is a machine learning model obtained by performing model training by taking historical operation behavior data and a processing result corresponding to the operation behavior data as training samples.
6. The risk prevention and control decision method of claim 5, wherein the operational behavior data comprises at least one of prior operational behavior prediction data, prior operational behavior data, and post-operational behavior data.
7. A risk prevention decision-making apparatus, comprising:
the risk event data determining unit is used for determining risk event data to be prevented and controlled;
the wind control strategy determining unit is used for inputting the risk event data to be prevented and controlled into a preset risk decision model to obtain a target wind control strategy output by the risk decision model;
the risk decision model is a machine learning model obtained by performing model training on the basis of sample risk event data and an actual wind control strategy corresponding to the sample risk event data;
the sample risk event data comprises historical risk event data and abnormal risk event data obtained through simulation based on the historical risk event data.
8. The risk prevention and control decision device of claim 7, further comprising: a sample pre-processing unit; the sample preprocessing unit is specifically configured to: acquiring initial sample risk event data; determining a target networking service terminal in the networking service terminals based on the user use state, the equipment operation state, the equipment performance parameters and the equipment load information of the networking service terminals; sending the initial sample risk event data to at least one corresponding target networking service terminal for preprocessing; and/or acquiring sample data obtained by locally acquiring data and preprocessing at least one target networking service terminal to obtain preprocessed sample data; and carrying out sample mixing on the preprocessed sample data, and selecting the sample risk event data meeting preset conditions in a sample fine adjustment and screening processing mode.
9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor when executing the computer program implements the steps of the risk prevention and control decision method of any one of claims 1 to 6.
10. A processor-readable storage medium, having stored thereon a computer program, wherein the computer program, when executed by a processor, performs the steps of a risk prevention decision method as claimed in any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210418916.3A CN114612011A (en) | 2022-04-20 | 2022-04-20 | Risk prevention and control decision method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210418916.3A CN114612011A (en) | 2022-04-20 | 2022-04-20 | Risk prevention and control decision method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114612011A true CN114612011A (en) | 2022-06-10 |
Family
ID=81869515
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210418916.3A Pending CN114612011A (en) | 2022-04-20 | 2022-04-20 | Risk prevention and control decision method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114612011A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115563657A (en) * | 2022-09-27 | 2023-01-03 | 冯淑芳 | Data information security processing method and system and cloud platform |
| CN115935265A (en) * | 2023-03-03 | 2023-04-07 | 支付宝(杭州)信息技术有限公司 | Method for training risk recognition model, risk recognition method and corresponding device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101498647B1 (en) * | 2014-01-21 | 2015-03-11 | (주)우산씨앤씨 | Security Management System And Security Management Method Using The Same |
| RU2570009C1 (en) * | 2014-08-06 | 2015-12-10 | Федеральное государственное унитарное предприятие "Центральный научно-исследовательский институт машиностроения" (ФГУП ЦНИИмаш) | Warning of danger in near-earth space and on earth and acs to this end |
| CN105574299A (en) * | 2016-02-19 | 2016-05-11 | 上海果路交通科技有限公司 | Safety pre-evaluation method for rail transit signal system |
| CN111325444A (en) * | 2020-01-21 | 2020-06-23 | 支付宝(杭州)信息技术有限公司 | Risk prevention and control decision method, device, system and equipment |
| CN112784420A (en) * | 2021-01-26 | 2021-05-11 | 支付宝(杭州)信息技术有限公司 | Simulation evaluation method, device and equipment for wind control strategy |
| CN113192328A (en) * | 2021-04-23 | 2021-07-30 | 长安大学 | Road operation risk prevention and control system and cooperative layout method of traffic signboard |
| CN113192327A (en) * | 2021-04-23 | 2021-07-30 | 长安大学 | Road operation risk active prevention and control system and method considering traffic flow and individuals |
| CN113610474A (en) * | 2021-08-16 | 2021-11-05 | 傲林科技有限公司 | Inventory management method and management system based on event network |
-
2022
- 2022-04-20 CN CN202210418916.3A patent/CN114612011A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101498647B1 (en) * | 2014-01-21 | 2015-03-11 | (주)우산씨앤씨 | Security Management System And Security Management Method Using The Same |
| RU2570009C1 (en) * | 2014-08-06 | 2015-12-10 | Федеральное государственное унитарное предприятие "Центральный научно-исследовательский институт машиностроения" (ФГУП ЦНИИмаш) | Warning of danger in near-earth space and on earth and acs to this end |
| CN105574299A (en) * | 2016-02-19 | 2016-05-11 | 上海果路交通科技有限公司 | Safety pre-evaluation method for rail transit signal system |
| CN111325444A (en) * | 2020-01-21 | 2020-06-23 | 支付宝(杭州)信息技术有限公司 | Risk prevention and control decision method, device, system and equipment |
| CN112784420A (en) * | 2021-01-26 | 2021-05-11 | 支付宝(杭州)信息技术有限公司 | Simulation evaluation method, device and equipment for wind control strategy |
| CN113192328A (en) * | 2021-04-23 | 2021-07-30 | 长安大学 | Road operation risk prevention and control system and cooperative layout method of traffic signboard |
| CN113192327A (en) * | 2021-04-23 | 2021-07-30 | 长安大学 | Road operation risk active prevention and control system and method considering traffic flow and individuals |
| CN113610474A (en) * | 2021-08-16 | 2021-11-05 | 傲林科技有限公司 | Inventory management method and management system based on event network |
Non-Patent Citations (2)
| Title |
|---|
| 邢巍等: "基于数据分析的业务风险防控研究", 现代商业, 28 March 2020 (2020-03-28), pages 15 - 17 * |
| 陈明等: "智能制造导论", vol. 1, 31 March 2021, 机械工业出版社, pages: 199 - 200 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115563657A (en) * | 2022-09-27 | 2023-01-03 | 冯淑芳 | Data information security processing method and system and cloud platform |
| CN115563657B (en) * | 2022-09-27 | 2023-12-01 | 国信金宏(成都)检验检测技术研究院有限责任公司 | Data information security processing method, system and cloud platform |
| CN115935265A (en) * | 2023-03-03 | 2023-04-07 | 支付宝(杭州)信息技术有限公司 | Method for training risk recognition model, risk recognition method and corresponding device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114612011A (en) | Risk prevention and control decision method and device | |
| CN112199581B (en) | Cloud computing and information security oriented cloud service management method and artificial intelligence platform | |
| CN112837069B (en) | Block chain and big data based secure payment method and cloud platform system | |
| CN112633962B (en) | Service recommendation method and device, computer equipment and storage medium | |
| CN110276369B (en) | Feature selection method, device and equipment based on machine learning and storage medium | |
| CN113435998B (en) | Loan overdue prediction method and device, electronic equipment and storage medium | |
| CN111582341B (en) | User abnormal operation prediction method and device | |
| CN115130536A (en) | Training method of feature extraction model, data processing method, device and equipment | |
| CN113205403A (en) | Method and device for calculating enterprise credit level, storage medium and terminal | |
| CN116109139A (en) | Wind control strategy generation method, decision method, server and storage medium | |
| CN112884569A (en) | Credit assessment model training method, device and equipment | |
| CN112766402A (en) | Algorithm selection method and device and electronic equipment | |
| CN111738824A (en) | Method, device and system for screening financial data processing modes | |
| CN116823428A (en) | Anti-fraud detection method, device, equipment and storage medium | |
| CN116915710A (en) | Traffic early warning method, device, equipment and readable storage medium | |
| CN118134288A (en) | Method and system for describing flexibility of clustered electric vehicle based on Gaussian model | |
| CN110704614B (en) | Information processing method and device for predicting user group type in application | |
| CN111598275B (en) | Electric vehicle credit score evaluation method, device, equipment and medium | |
| CN110855474B (en) | Network feature extraction method, device, equipment and storage medium of KQI data | |
| CN117196630A (en) | Transaction risk prediction method, device, terminal equipment and storage medium | |
| CN115713669A (en) | Image classification method and device based on inter-class relation, storage medium and terminal | |
| CN115631457A (en) | Man-machine cooperation abnormity detection method and system in building construction monitoring video | |
| CN111654853B (en) | Data analysis method based on user information | |
| CN111860661B (en) | Data analysis method and device based on user behaviors, electronic equipment and medium | |
| CN116451050A (en) | Abnormal behavior recognition model training and abnormal behavior recognition method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |