WO2021203489A1 - 一种记录决策行为的方法、系统和设备 - Google Patents

一种记录决策行为的方法、系统和设备 Download PDF

Info

Publication number
WO2021203489A1
WO2021203489A1 PCT/CN2020/086253 CN2020086253W WO2021203489A1 WO 2021203489 A1 WO2021203489 A1 WO 2021203489A1 CN 2020086253 W CN2020086253 W CN 2020086253W WO 2021203489 A1 WO2021203489 A1 WO 2021203489A1
Authority
WO
WIPO (PCT)
Prior art keywords
decision
record model
record
model
making
Prior art date
Application number
PCT/CN2020/086253
Other languages
English (en)
French (fr)
Inventor
刘煜
孙再连
梅瑜
Original Assignee
厦门邑通软件科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 厦门邑通软件科技有限公司 filed Critical 厦门邑通软件科技有限公司
Publication of WO2021203489A1 publication Critical patent/WO2021203489A1/zh

Links

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
    • G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
    • 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
    • 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
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/02—Banking, e.g. interest calculation or account maintenance
    • 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
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/03—Credit; Loans; Processing thereof
    • 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
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • 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
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • 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
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • G06Q50/205—Education administration or guidance
    • 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
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/20—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms

Definitions

  • the present invention relates to the technical field of decision management, in particular to a method, system and equipment for recording decision behaviors.
  • the process flow of production lines or equipment in the industrial field has been relatively solidified, and the basic working condition data dimensions on which decision-making or operation behaviors are based, the data dimensions that workers can operate, and the evaluation data dimensions for evaluating operation results are all clear.
  • decision-making events such as community service, social security, government department strategies, etc.
  • the same or the same type of decision-making event generally has the same decision-making purpose, evaluation data dimension, and evaluation standard, but different people
  • the data dimensions on which decision-making is based may be different, and the corresponding decision-making behaviors (ie, operating data dimensions) are also diverse.
  • the embodiments of the present invention provide a method, system, and equipment for recording decision-making behaviors.
  • the core of the method is to obtain a method for recording decision-making behaviors, which can effectively record decision-making behaviors in some special scenarios and reduce the accumulation and inheritance of decision-making experience.
  • the difficulty is to obtain a method for recording decision-making behaviors, which can effectively record decision-making behaviors in some special scenarios and reduce the accumulation and inheritance of decision-making experience.
  • an embodiment of the present invention provides a method for recording a decision-making behavior, the method including:
  • the different decision record models perform independent learning on the decision event data set and grow together within a set time
  • the decision-making record model cluster is assessed, and the decision-making record model is managed according to the set elimination rules.
  • an embodiment of the present invention provides a system for recording decision-making behaviors, the system including:
  • the model building module is configured to input or build a decision record model to form a decision record model cluster
  • the data collection module is configured to collect or input decision-making event data sets for learning by the decision-making record model
  • the model analysis module is configured to input decision events to evaluate decision record model clusters and eliminate inferior decision record models
  • the prediction module is configured to input decision events to output operation decisions and result predictions.
  • an embodiment of the present invention provides a device for recording decision-making behavior.
  • the device includes a memory and a processor; wherein executable code is stored in the memory, and when the executable code is executed by the processor When the time, the processor can at least implement the method of recording decision-making behavior in the first aspect.
  • the embodiment of the present invention also provides a non-transitory machine-readable storage medium having executable code stored on the non-transitory machine-readable storage medium, and when the executable code is executed by a processor of an electronic device, The processor can at least implement the method of recording decision-making behavior in the first aspect.
  • a method of recording decision-making behaviors which facilitates the recording, accumulation, selection and inheritance of decision-making events in a group, and the present invention adopts a clustering co-growth mechanism, taking into account the decision-making record model
  • the personalized development and the enhancement of common value can be oriented to a wide range of application scenarios such as service industry, regional governance, social security, government department decision-making, etc., has wide applicability and high practicability, and provides a method for constructing large-scale, comprehensive intelligent decision-making scenarios for various scenarios.
  • FIG. 1 is a flowchart of a method for recording decision-making behaviors according to an embodiment of the present invention
  • Figure 2 is a block diagram of a system provided by an embodiment of the present invention.
  • Fig. 3 is a schematic structural diagram of a medium according to an embodiment of the present invention.
  • FIG. 4 is a schematic structural diagram of a device provided by an embodiment of the present invention.
  • the words “if” and “if” as used herein can be interpreted as “when” or “when” or “in response to determination” or “in response to detection”.
  • the phrase “if determined” or “if detected (statement or event)” can be interpreted as “when determined” or “in response to determination” or “when detected (statement or event) )” or “in response to detection (statement or event)”.
  • decision-making events such as community service, social security, government department strategies, etc.
  • the same or the same type of decision-making event generally has the same decision-making purpose, evaluation data dimension, and evaluation standard, but different people are making decisions.
  • the time-based data dimensions may be different, and the corresponding decision-making behaviors (ie, operating data dimensions) are also diverse.
  • the present invention provides a method, system and equipment for recording decision-making behaviors.
  • multiple decision-making record models grow together, and then through the elimination of inferior decision-making record models, so as to realize the update, retain the decision-making record model that conforms to the decision-making event, and realize the learning and experience of the decision-making event. Accumulate and make predictions for future decision-making events.
  • the implementation principles of the methods, systems, media, and equipment are similar, and will not be repeated here.
  • the embodiments of the present invention can be applied to various scenarios and various device types to record decision-making behaviors of decision-making events. It should be noted that the embodiments provided by the present invention are only shown to facilitate the understanding of the spirit and principle of the present invention, and the embodiments of the present invention are not limited in this respect. On the contrary, the embodiments of the present invention can be applied to any applicable scenarios.
  • the embodiment of the present invention provides a method for recording decision-making behavior. As shown in FIG. 1, the method includes:
  • the different decision record models perform independent learning on the decision event data set, and grow together within a set time;
  • it also includes S104, which predicts the decision strategy and the strategy result of the decision event through the decision record model cluster.
  • the decision event data set includes a decision event
  • the decision event includes a dependency data dimension, an operation data dimension, and an evaluation data dimension.
  • the decision recording model After the data is discretized, the decision recording model records X, Y, and Z.
  • the evaluation data dimensions of the decision events in the same scene are the same, that is, the evaluation method of the decision result is good or bad, or the expression is the same, that is, a group of decision record models with the same evaluation data Z constitutes a decision record model cluster .
  • the decision record model cluster of the present invention contains multiple decision record models and can eliminate inferior decision record models through a competitive survival mechanism
  • the decision record model cluster has a fault-tolerant mechanism.
  • the decision record model is constructed, that is, in S101, the user Or user groups can build their own decision-making record models based on their own habits or the summary of actual decision-making experience, without requiring these decision-making record models to be reasonable and effective.
  • the aforementioned competitive survival mechanism requires that decision-making recording models need to grow together for a period of time. During this time, by learning the decision-making event data set, one decision-making event can match multiple decision-making recording models. These decision-making recording models independently record and learn decision-making events. Perfect the model itself.
  • the decision record model is processed according to the elimination rules, and the elimination rules are not limited to the following:
  • the average value of the prediction results of a decision record model refers to the average value of the expression value of Z of the existing decision event records of the decision record model.
  • decision record model a contains decision record model b.
  • decision record model a When decision record model a is significantly better than decision record model b, delete decision record model b, which is called decision record model a which absorbs decision record model b ;
  • decision record model b When the decision record model b is significantly better than the decision record model a, delete the decision record model a, which is called the decision record model b instead of the decision record model a;
  • the sculpting of the decision record model refers to deleting several data dimensions in ⁇ X ⁇ of a decision record model, or deleting several data dimensions in ⁇ Y ⁇ to generate a new and simpler decision record model.
  • each independent decision record model can accumulate more decision-making event records after a period of coexistence and coexistence.
  • the present invention provides a competitive survival mechanism for the decision-making behavior decision record model cluster. , So that the decision record model in the decision record model cluster can achieve the goal of eliminating the inferior and the good growing and surviving and evolving to be better.
  • new decision record models can be continuously added to keep the decision record model cluster constantly evolving and updating.
  • Scenario 1 Decision-making event of enterprise micro-loan placement
  • S101 Generate a decision record model cluster, and generate a decision record model
  • S102 Different decision recording models perform independent learning on the decision event data set and grow together within a set time. For each loan and repayment business, a matching decision record model is triggered to generate a decision behavior record.
  • S103 Periodic assessment, with a one-year cycle as the final elimination of decision-making record models, but the number of decision-making record models is maintained at 3 or more.
  • S101 Generate a decision record model cluster, and generate a decision record model
  • ⁇ Z ⁇ ⁇ number of people receiving this service, average service target score, community leader’s score ⁇
  • its expression proportion of service targets*25+number of people receiving this service/(proportion of service targets* The total population of the service area)*20+the average satisfaction value of the service object+the community leader’s score; direction: the higher the better.
  • ⁇ X ⁇ includes at least: the service area, the total population of the service area, the type of service object, the proportion of the service object, and the annual per capita cost of the service object;
  • ⁇ Y ⁇ contains at least: the purpose of the service, the form of the service, and the content of the service.
  • S102 Different decision-making recording models perform independent learning on the decision-making event data set, and grow together within a set time. Each time the community serves, the matching decision record model is triggered to generate decision behavior records.
  • S103 Periodic assessment, calculate the average value of the evaluation data of each decision record model, eliminate the average value of the evaluation data below the set value, sort the average value of the evaluation data of the decision record model from high to bottom, and determine the priority of the procurement service in the second year .
  • S101 Generate a decision record model cluster, and generate a decision record model
  • ⁇ X ⁇ includes at least: the distribution of reserve points, the reserve amount of emergency materials, the transportation time of emergency materials, the material requirements of the emergency points, the restriction period, the emergency degree/emergency event level;
  • ⁇ Y ⁇ contains at least: emergency material delivery plan, including emergency material delivery type, emergency material quantity, storage point selection, and delivery point selection.
  • S102 Different decision recording models perform independent learning on the decision event data set and grow together within a set time. For each emergency event and emergency material demand, the matching decision record model is triggered to generate decision behavior records.
  • S103 Periodic assessment, calculating the evaluation data of each decision record model, sorting the performance of the evaluation data from low to high, and determining the priority of the next emergency material delivery program selection model.
  • S101 Generate a decision record model cluster, and generate a decision record model
  • Each college configures ⁇ X ⁇ , ⁇ Y ⁇ according to their respective educational resource allocation plan:
  • ⁇ X ⁇ At least include: teaching venues and teachers;
  • ⁇ Y ⁇ contains at least: course offering plan.
  • S103 Periodic assessment, sorting according to the current Z value, and eliminating the decision record model at the last place.
  • S101 Generate a decision record model cluster, and generate a decision record model
  • ⁇ X ⁇ contains at least: the current number of various medical devices and the number of patients;
  • ⁇ Y ⁇ includes at least: the type of imported medical devices, the number of imported medical devices, and the charging standards.
  • S102 Different decision recording models perform independent learning on the decision event data set and grow together within a set time. Each medical device investment demand triggers a matching decision record model to generate decision behavior records.
  • S103 Periodic assessment, sorting according to the average value of Z value of the current period, and making the final elimination of the decision record model.
  • S101 Generate a decision record model cluster, and generate a decision record model
  • the enterprise has a total of 200 salespersons. According to personal experience, the salespersons rely on data dimensions ⁇ X ⁇ and configuration operation data ⁇ Y ⁇ for configuration decisions:
  • ⁇ Y ⁇ contains at least: investment amount.
  • S102 Different decision recording models perform independent learning on the decision event data set and grow together within a set time. Each time an investment decision is made, a matching decision record model is triggered to generate a decision behavior record.
  • S103 Periodic assessment, sorting according to the average value of Z value of the current period, and eliminating the final decision record model.
  • Scenario 7 Decision-making event of increasing parking spaces on urban streets
  • S101 Generate a decision record model cluster, and generate a decision record model
  • Y includes at least: charging standards.
  • S102 Different decision recording models perform independent learning on the decision event data set and grow together within a set time. Every time a parking space is charged, a matching decision record model is triggered to generate a decision behavior record.
  • S103 Periodic assessment, sorting according to the current Z value, and eliminating the final decision record model.
  • S101 Generate a decision record model cluster, and generate a decision record model
  • the enrollment areas are divided by city and county, and the year before the implementation of this decision model is used as the reference period. Each area can adopt different enrollment strategies.
  • Each enrollment season provides regional allocation ⁇ X ⁇ , ⁇ Y ⁇ :
  • ⁇ X ⁇ includes at least the area of enrollment, the number of students enrolled in the regional reference period, the per capita cost of enrollment in the regional reference period, the set of enrollment majors, the size of the regional economy, and the size of the regional source of students;
  • ⁇ Y ⁇ At least include: admissions promotion channels, institution introduction key points, professional introduction sequence, professional introduction key points.
  • S102 Different decision recording models perform independent learning on the decision event data set and grow together within a set time. Each enrollment area in each enrollment season triggers a matching decision-making record model to generate decision-making behavior records.
  • S103 Periodic assessment.
  • the decision record model is eliminated at the end of the enrollment season, but the number of decision record models is maintained at 2 or more; after the decision record model of a region is eliminated, the next enrollment season can choose to remain Or create a new decision record model.
  • S101 Generate a decision record model cluster, and generate a decision record model
  • ⁇ X ⁇ contains at least: regional population, regional area, regional GDP in the previous year, and regional economic scale;
  • ⁇ Y ⁇ contains at least: economic development plan, key support industries, support plans, total support funds, energy conservation and emission reduction measures, investment promotion measures, and talent introduction measures.
  • S103 Periodic assessment, with a one-year cycle as the final elimination of decision-making record models, but the number of decision-making record models is maintained at 3 or more. After the decision-making record model of a city or county is eliminated, the remaining decision-making record model will be given priority in the next year or a new decision-making record model will be created.
  • S101 Generate a decision record model cluster, and generate a decision record model
  • ⁇ X ⁇ includes at least: battlefield situation, enemy targets, enemy target types, strategy ⁇ campaign targets, deployable forces, deployable weapons and ammunition bases, maneuverability, and theater supply capabilities;
  • ⁇ Y ⁇ contains at least: the use of troops, the weapons used, the quantity, the maneuvering route, the post-war transfer plan, and the replenishment strategy.
  • S102 Different decision-making recording models perform independent learning on the decision-making event data set, and grow together within a set time. Each exercise or actual combat triggers the matching decision-making record model to generate decision-making behavior records.
  • S103 Periodic assessment, with a one-year cycle as the final elimination of decision-making record models, but the number of decision-making record models is maintained at 3 or more.
  • S101 Generate a decision record model cluster, and generate a decision record model
  • the human resources administration department of the enterprise configures ⁇ X ⁇ , ⁇ Y ⁇ according to the elements:
  • S101 Generate a decision record model cluster, and generate a decision record model
  • the staff of the mall's investment promotion department are configured according to the elements ⁇ X ⁇ , ⁇ Y ⁇ :
  • ⁇ X ⁇ contains at least: consumer groups
  • ⁇ Y ⁇ includes at least: publicity channels, investment strategy, and pricing strategy.
  • the present invention provides a system for recording decision-making behaviors, which can implement the method for recording decision-making behaviors in the exemplary embodiment of the present invention corresponding to FIG. 1.
  • the system includes: model building module, data acquisition module, model analysis module, and prediction module. Specifically:
  • the model building module is configured to input or build a decision record model to form a decision record model cluster
  • the data collection module is configured to collect or input decision-making event data sets for learning by the decision-making record model
  • the model analysis module is configured to input decision events to evaluate decision record model clusters and eliminate inferior decision record models
  • the prediction module is configured to input decision events to output operation decisions and result predictions.
  • the present invention provides an exemplary medium storing computer-executable instructions, and the computer-executable instructions can be used to make all
  • the computer executes the method of the example of the present invention corresponding to FIG. 1.
  • the device 40 includes a processing unit 401, a memory 402, a bus 403, An external device 404, an I/O interface 405, and a network adapter 406.
  • the memory 402 includes a random access memory (RAM) 4021, a cache memory 4022, a read-only memory (Read-Only Memory, ROM) 4023, and at least A memory cell array 4025 composed of a memory cell 4024.
  • the memory 402 is used to store programs or instructions executed by the processing unit 401; the processing unit 401 is used to execute the method according to the example of the present invention corresponding to FIG. 1 according to the programs or instructions stored in the memory 402; the I/ The O interface 405 is used to receive or send data under the control of the processing unit 401.
  • the exemplary device 40 includes, but is not limited to, user equipment, network equipment, or a device formed by integrating network equipment and user equipment through a network;
  • the user equipment includes, but is not limited to, any type that can communicate with the user through a keyboard.
  • Remote control, touchpad or voice control equipment for human-computer interaction electronic products such as computers, smart phones, ordinary mobile phones, tablet computers, etc.
  • the network equipment includes but not limited to computers, network hosts, a single network server, multiple networks A set of servers or a cloud composed of multiple servers.
  • each implementation manner can be implemented by adding a necessary general hardware platform, and of course, it can also be implemented by a combination of hardware and software.
  • the above technical solution essentially or the part that contributes to the prior art can be embodied in the form of a computer product, and the present invention can be used in one or more computer usable storage containing computer usable program codes.
  • the form of a computer program product implemented on a medium including but not limited to disk storage, CD-ROM, optical storage, etc.).

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Educational Administration (AREA)
  • Primary Health Care (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Educational Technology (AREA)
  • Public Health (AREA)
  • Medical Informatics (AREA)
  • Epidemiology (AREA)
  • Biomedical Technology (AREA)
  • Technology Law (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种记录决策行为的方法、系统和设备，该方法包括：构建决策记录模型簇，所述决策记录模型簇包括不同的决策记录模型（S101）；所述不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长（S102）；对所述决策记录模型簇考核，根据设定的淘汰规则对决策记录模型做更新（S103）。该方法方便了群体中决策事件的记录、积累、选优和传承，且采用丛生共长机制，兼顾了决策记录模型的个性化发展和共性价值提升，具有广泛的应用场景，适用性广，实用性高。

Description

一种记录决策行为的方法、系统和设备 技术领域
本发明涉及决策管理技术领域，尤其涉及一种记录决策行为的方法、系统和设备。
背景技术
一般情况下，工业领域的生产线或设备，其工艺流程已经相对固化，决策行为或操作行为所依据的基础工况数据维度、工人可操作的数据维度以及评价操作结果的评价数据维度都是明确的。但在一些决策事件中，如社区服务、社会治安、政府部门策略等场景，同一个或同一种类型的决策事件，一般具有相同的决策目的、评价数据维度、评价标准，只不过，不同人在决策时所依据的数据维度可能是不同的，其相应的决策行为(即操作数据维度)也是多样的，虽然每个人都可以总结经验教训，但经验之间的可比性差，经验的积累和传承困难大，因此迫切需要解决决策经验的积累以及决策经验在人与人、人与系统、系统与人之间的转移问题。
发明内容
本发明实施例提供一种记录决策行为的方法、系统和设备，其核心在于得出一种记录决策行为的方法，能够有效对一些特殊场景的决策行为进行记录，减小决策经验的积累与传承的难度。
第一方面，本发明实施例提供一种记录决策行为的方法，该方法包括：
构建决策记录模型簇，所述决策记录模型簇包括不同的决策记录模型；
所述不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内 共同生长；
对所述决策记录模型簇考核，根据设定的淘汰规则对决策记录模型做管理。
第二方面，本发明实施例提供一种记录决策行为的系统，该系统包括：
模型建立模块，被配置为用于输入或建立决策记录模型，以形成决策记录模型簇；
数据采集模块，被配置为用于采集或输入决策事件数据集，以供决策记录模型学习；
模型分析模块，被配置为输入决策事件以考核决策记录模型簇，并淘汰劣势决策记录模型；
预测模块，被配置为输入决策事件以输出操作决策和结果预测。
通过该系统，至少可以实现第一方面中的记录决策行为的方法
第三方面，本发明实施例提供一种记录决策行为的设备，该设备包括：存储器、处理器；其中，所述存储器上存储有可执行代码，当所述可执行代码被所述处理器执行时，使所述处理器至少可以实现第一方面中的记录决策行为的方法。
本发明实施例还提供了一种非暂时性机器可读存储介质，所述非暂时性机器可读存储介质上存储有可执行代码，当所述可执行代码被电子设备的处理器执行时，使所述处理器至少可以实现第一方面中的记录决策行为的方法。
在本发明实施例中，给出了一种记录决策行为的方法，该方法方便了群体中决策事件的记录、积累、选优和传承，且本发明采用丛生共长机制，兼顾了决策记录模型的个性化发展和共性价值提升。本发明能够面向服务业、区域治理、社会治安、政府部门决策等广泛的应用场景，适用性广，实用性高，为各种场景提供了构建大型、综合智慧决策场景的方法。
附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图是本发明的 一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本发明一实施例提供的一种记录决策行为的方法的流程图；
图2为本发明一实施例提供的一种系统的模块图；
图3本本发明一实施例提供的一种介质的结构示意图；
图4为本发明一实施例提供的一种设备的结构示意图；
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
在本发明实施例中使用的术语是仅仅出于描述特定实施例的目的，而非旨在限制本发明。在本发明实施例和所附权利要求书中所使用的单数形式的“一种”、“所述”和“该”也旨在包括多数形式，除非上下文清楚地表示其他含义，“多种”一般包含至少两种。
取决于语境，如在此所使用的词语“如果”、“若”可以被解释成为“在……时”或“当……时”或“响应于确定”或“响应于检测”。类似地，取决于语境，短语“如果确定”或“如果检测(陈述的条件或事件)”可以被解释成为“当确定时”或“响应于确定”或“当检测(陈述的条件或事件)时”或“响应于检测(陈述的条件或事件)”。
在一些决策事件中，如社区服务、社会治安、政府部门策略等场景，同一个或同一种类型的决策事件，一般具有相同的决策目的、评价数据维度、评价标准，只不过，不同人在决策时依据的数据维度可能是不同的，其相应的决策行为(即操作数据维度)也是多样的，虽然每个人都可以总结经验教训，但经验之间的可比性差，经验的积累和传承困难大，因此迫切需要解决决策经验的 积累以及决策经验在人与人、人与系统、系统与人之间的转移问题。
针对上述问题，本发明提供了一种记录决策行为的方法、系统和设备。本发明提供的技术方案中，主要通过让多个决策记录模型共同生长，再通过淘汰劣势的决策记录模型，从而实现更新，保留符合决策事件的决策记录模型，实现对决策事件的经验的学习和积累，为往后的决策事件做出预测。其中，方法、系统、介质和设备的实现原理相似，此处不再赘述。
在介绍了本发明的基本原理之后，下面具体介绍本发明的各种非限制性实施方式。
本发明实施例可以应用于各种场景和各种设备类型，用以记录决策事件的决策行为。需要注意的是，本发明提供的实施例仅是为了便于理解本发明的精神和原理而示出，本发明的实施方式在此方面不受任何限制。相反，本发明的实施方式可以应用于适用的任何场景。
本发明实施例提供了一种记录决策行为的方法，如图1所示，该方法包括：
S101、构建决策记录模型簇，所述决策记录模型簇包括不同的决策记录模型；
S102、所述不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长；
S103、对所述决策记录模型簇考核，根据设定的淘汰规则对决策记录模型做管理，进而更新决策记录模型簇；
在其它实施例中，还有包括S104、通过决策记录模型簇对决策事件的决策策略和策略结果进行预测。
本实施例中，所述决策事件数据集包括决策事件，所述决策事件包括依赖数据维度、操作数据维度和评价数据维度，具体的，
所述依赖数据维度包括决策事件所依赖的当前事实，事务状况的数据维度，记为X，X＝{X}＝{X 1，X 2，X 3...，X a}，其中，X i为第i个依赖数据，1≤i≤a，a是依赖数据的维度数；
所述操作数据维度包括决策事件的决策内容、决策后行为的数据维度，记为Y，Y＝{Y}＝{Y 1，Y 2，Y 3...，Y b}，其中，Y i为第i个操作数据，1≤i≤b，b是操作数据的维度数；
所述评价数据维度包括决策事件的决策结果优劣评价指标、方式、公式，记为Z，Z＝{Z}＝{Z 1，Z 2，Z 3...，Z c}，其中，Z i为第i个评价数据，1≤i≤c，c是评价数据的维度数。
数据离散化后，所述决策记录模型对X，Y，Z进行记录。
需要注意的是，同一场景的决策事件的评价数据维度相同，即决策结果优劣的评价方式，或者表达式是一样的，即一组具有相同评价数据Z的决策记录模型构成一个决策记录模型簇。
由于本发明的决策记录模型簇包含多个决策记录模型，且能够通过竞争生存机制淘汰劣质的决策记录模型，所以决策记录模型簇具备容错机制，在构建决策记录模型时，即在S101中，用户或者用户群可以按照自身习惯或者实际决策经验的总结构建自己的决策记录模型，而不要求这些决策记录模型必须是合理的，有效的。
上述的竞争生存机制要求决策记录模型需要共同生长一段时间，在这段时间内，通过学习决策事件数据集，一个决策事件可以匹配多个决策记录模型，这些决策记录模型独立记录并学习决策事件，完善模型自身。
决策记录模型共同生长一段时间后，需要进行考核，淘汰劣势的决策记录模型，即进入S103。
考核过程中，在决策记录模型簇的各自独立的决策记录模型里输入相同决策事件各自的X值，输出各自历史最优或计算出的的Y值和相应的Z值，并建立相应的淘汰规则。
考核中，根据淘汰规则对决策记录模型进行处理，淘汰规则不限于下述几种：
依据一段时间内的Z值平均值进行排序，对决策记录模型末位淘汰；
对具有包含关系的决策记录模型之间，通过吸收或取代方式淘汰劣势决策 记录模型；
通过合并决策记录模型生成新的决策记录模型；
通过模型塑身生成更简单的决策记录模型。
具体的：
一个决策记录模型的预测结果的平均值是指，该决策记录模型现有的各决策事件记录的Z的表达式值的平均值。
在一个决策记录模型簇内，决策记录模型a包含决策记录模型b，当决策记录模型a明显优于决策记录模型b时，删除决策记录模型b，称为决策记录模型a吸收了决策记录模型b；当决策记录模型b明显优于决策记录模型a时，删除决策记录模型a，称为决策记录模型b取代了决策记录模型a；
令决策记录模型a和决策记录模型b是一个决策记录模型簇内的两个决策记录模型，称决策记录模型c是a、b的并,当且仅当{X} c＝{X} a∪{X} b，{Y} c＝{Y} a∪{Y} b，{Z} c＝{Z} a＝{Z} b；
决策记录模型塑身是指删除一个决策记录模型的{X}中的若干个数据维度，或删除{Y}中的若干个数据维度，生成新的更简单的决策记录模型。
以下是决策记录模型之间优劣性的判断方式：
决策记录模型a和决策记录模型b是一个决策记录模型簇中的两个决策记录模型,v＝a的Z的均值-b的Z的均值,
当v>0且Z的方向性是越大越优时，称a优于b或b劣于a；
当v>0且Z的方向性是越小越优时，称b优于a或a劣于b；
当v>预设的显著值>0且Z的方向性是越大越优时，称a明显优于b或b明显劣于a；
当v>预设的显著值>0且Z的方向性是越小越优时，称b明显优于a或a明显劣于b。
显然，本发明决策记录模型簇内，各个独立的决策记录模型经过一段时间的同生共长，可以积累更多的决策事件记录，同时，本发明为决策行为决策记 录模型簇提供了竞争生存机制，使决策记录模型簇中的决策记录模型得以实现劣者被淘汰，优者成长生存、进化更优的目标。
在其它实施例中，淘汰掉一部分的决策记录模型后，还可以不断的加入新的决策记录模型，保持决策记录模型簇的不断进化更新。
在介绍了本发明示例性实施方式的方法之后，接下来，将上述实施例应用于具体的场景中。
场景一：企业小额贷款投放决策事件
上述方法的具体实施方式如下：
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度{Z}＝{还贷逾期天数}，方向：越小越优。
企业共有员工200人，由员工根据个人经验，配置决策依赖数据维度{X}，配置共同的操作数据{Y}＝{贷款金额，贷款期限}。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。每次贷款、还款业务，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，以一年为周期对决策记录模型做末位淘汰，但决策记录模型数量维持在大于等于3个。
场景二：社区服务决策
S101：生成决策记录模型簇，生成决策记录模型；
具体的，{Z}＝{接受本次服务的人数，服务对象评分均值，社区分管领导评分}，其表达式＝服务对象占比*25+接受本次服务的人数/(服务对象占比*服务区域总人口)*20+服务对象满意度均值+社区分管领导评分；方向：越高越优。
当年各社区服务提供团体配置{X}、{Y}：
{X}至少包含：服务的区域，服务区域总人口，服务对象类型，服务对象占比，服务对象年人均费用；
{Y}至少包含：服务的宗旨、服务的形式、服务的内容。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间 内共同生长。每次社区服务，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，计算各决策记录模型的评价数据均值，对评价数据均值低于设定值的做淘汰，按决策记录模型的评价数据均值从高到底排序，确定第二年采购服务的优先级。
场景三：城市应急物资投放决策事件
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度{Z}＝{满足需求的应急物资调度时长}，方向：越小越优。
相关省级、市级政府应急部门和机构，如应急管理办公室配置{X}、{Y}：
{X}至少包含：储备点的分布、应急物资的储备量、应急物资的运输时长、应急点的物资需求量、限制期、应急程度/应急事件级别；
{Y}至少包含：应急物资投放方案，包括应急物资投放种类、应急物资投放数量、储备点选择、投放点选择。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。每次应急事件、应急物资需求，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，计算各决策记录模型的评价数据，对评价数据表现从低到高排序，确定下一次应急物资投放方案选择模型的优先级。
场景四：教育资源分配决策
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度{Z}＝{招生增长率}，方向：越高越优。
各学院根据各自教育资源分配方案配置{X}、{Y}：
{X}至少包含：教学场所、师资情况；
{Y}至少包含：课程开设方案。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。每学期课程开设需求，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，依据当期的Z值进行排序，对决策记录模型做末位淘汰。
场景五：医疗资源投资决策
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度{Z}＝{医疗器械投资回报率}，方向：越高越优。
医院科室采购决策人根据当前医院各类器械使用情况及下一年度的利润目标配置{X}、{Y}：
{X}至少包含：当前各类医疗器械数量、患者数量；
{Y}至少包含：引进医疗器械类别、引进医疗器械数量、收费标准。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。每次医疗器械投资需求，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，依据当期的Z值平均值进行排序，对决策记录模型做末位淘汰。
场景六：企业投资决策事件
上述方法的具体实施方式如下：
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度{Z}＝{投资回报率}，方向：越高越优。
企业共有业务员200人，由业务员根据个人经验，配置决策依赖数据维度{X}、配置操作数据{Y}：
{Y}至少包含：投资金额。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。每次投资决策，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，根据当期的Z值平均值进行排序，对决策记录模型末位淘汰。
场景七：城市街道增加停车位决策事件
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度{Z}＝{地段车位数，车位收费收入}，其表达式＝ 车位收费收入/地段车位数，方向：越高越优。
当地街道工作人员根据不同地段、不同时长的收费配置决策依赖数据维度{X}，配置操作数据{Y}：
Y至少包含：收费标准。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。每次车位收费，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，根据当期的Z值进行排序，对决策记录模型末位淘汰。
场景八：高校招生策略决策
S101：生成决策记录模型簇，生成决策记录模型；
具体的，{Z}＝{区域实际招收人数，区域参照期招收人数，区域招生人均成本，区域参照期招生人均成本}，其表达式＝区域实际招收人数/区域参照期招收人数*100*5+(区域参照期招生人均成本-区域招生人均成本)*20+区域实际招收人数/区域生源规模*100；方向：越高越优。
按市、县划分招生区域，以本决策模式实施前一年为参照期，各区域可以采取不同的招生策略，各招生季提供区域配置{X}、{Y}：
{X}至少包含：招生的区域，区域参照期招生人数，区域参照期招生人均成本，招生专业集，区域经济规模，区域生源规模；
{Y}至少包含：招生宣传渠道、院校介绍重点、专业介绍次序、专业介绍重点。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。各招生区域每个招生季，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，以一个招生季为周期对决策记录模型做末位淘汰，但决策记录模型数量维持在大于等于2个；一个区域的决策记录模型被淘汰后，下一个招生季可选择剩存的决策记录模型或创建新决策记录模型。
场景九：经济未来规划走势决策
上述方法的具体实施方式如下：
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度{Z}＝{区域GDP，GDP每个百分点的能耗，污染指数}，其表达式＝各区域平均GDP*10000-GDP每个百分点的能耗(百万千瓦时)-污染指数*50；方向：越大越优。
辖区内共13个区县，各区县根据情况建立经济发展决策记录模型，配置{X}、{Y}：
{X}至少包含：区域人口，区域面积，区域上年GDP，区域经济规模；
{Y}至少包含：经济发展计划，重点扶持行业、扶持方案、扶持资金总量，节能减排措施，招商引资措施，人才引进措施。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。每季度触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，以一年为周期对决策记录模型做末位淘汰，但决策记录模型数量维持在大于等于3个。一个市县的决策记录模型被淘汰后，下一年度优先选择剩存的决策记录模型或创建新决策记录模型。
场景十：军事资源调度布局决策
上述方法的具体实施方式如下：
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度Z＝{目标毁损等级，我方是否暴露，消耗成本等级，补充难度等级}，其表达式＝目标毁损等级(彻低毁损：0，丧失战力：20，战力减半：50，战力未减：100)+我方是否暴露(是：20，否：0分)+消耗成本等级+补充难度等级；方向：越小越优。
参谋人员根据各战法要素配置决策记录模型的{X}、{Y}：
{X}至少包含：战场态势，敌方目标，敌方目标类型，战略\战役目标，可部署的兵力，可部署兵力的兵器及弹药基数、机动能力，战区补给能力；
{Y}至少包含：动用兵力，使用的兵器、数量，机动路线，战后转移方案，补给策略。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间 内共同生长。每次演习或实战触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，以一年为周期对决策记录模型做末位淘汰，但决策记录模型数量维持在大于等于3个。
场景十一：人力行政部招聘决策事件
上述方法的具体实施方式如下：
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度{Z}＝{愿景得分，身份认同得分，信念得分，能力得分，行为得分，知识得分}，其表达式＝愿景得分*30％+身份认同得分*20％+信念得分*20％+能力得分*10％+行为得分*10％+知识得分*10％；方向：越高越优。
企业人力行政部根据要素配置{X}、{Y}：
配置共同的操作数据{Y}＝{招聘}。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。每次招聘，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，以一年为周期对上一年度入职员工做考核，根据员工的考核结果对招聘模型做末位淘汰。
场景十二：商场招商决策事件
上述方法的具体实施方式如下：
S101：生成决策记录模型簇，生成决策记录模型；
具体的，设置评价数据维度{Z}＝{商场面积出租率}；方向：越高越优。
商场招商部工作人员根据要素配置{X}、{Y}：
{X}至少包含：消费群体；
{Y}至少包含：宣传渠道、招商策略、定价策略。
S102：不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长。每次招商需求，触发匹配的决策记录模型生成决策行为记录。
S103：定期考核，以一年为周期对决策记录模型做末位淘汰。
在介绍了本发明示例性实施方式的方法之后，接下来，介绍本发明提供了 示例性实施的系统。
请参阅图2，本发明提供了一种记录决策行为的系统，该系统可以实现图1对应的本发明示例性实施方式中的记录决策行为的方法。该系统包括：模型建立模块、数据采集模块、模型分析模块、预测模块，具体的：
模型建立模块，被配置为用于输入或建立决策记录模型，以形成决策记录模型簇；
数据采集模块，被配置为用于采集或输入决策事件数据集，以供决策记录模型学习；
模型分析模块，被配置为输入决策事件以考核决策记录模型簇，并淘汰劣势决策记录模型；
预测模块，被配置为输入决策事件以输出操作决策和结果预测。
本实施例的系统，其实现原理与方法的技术方案相似，此处不再赘述。
在介绍了本发明示例性实施方式的方法和装置之后，接下来，参考图3，本发明提供了一种示例性介质，该介质存储有计算机可执行指令，该计算机可执行指令可用于使所述计算机执行图1对应的本发明示例的方法。
在介绍了本发明示例性实施方式的方法、系统和介质之后，接下来，参考图4，介绍本发明提供的一种示例性设备40，该设备40包括处理单元401、存储器402、总线403、外部设备404、I/O接口405以及网络适配器406，该存储器402包括随机存取存储器(random access memory，RAM)4021、高速缓存存储器4022、只读存储器(Read-Only Memory，ROM)4023以及至少一片存储单元4024构成的存储单元阵列4025。其中该存储器402，用于存储处理单元401执行的程序或指令；该处理单元401，用于根据该存储器402存储的程序或指令，执行图1对应的本发明示例所述的方法；该I/O接口405，用于在该处理单元401的控制下接收或发送数据。
在此，所述示例性设备40其包括但不限于用户设备、网络设备或网络设备与用户设备通过网络相集成所构成的设备；所述用户设备包括但不限于任何一种可与用户通过键盘、遥控器、触摸板或声控设备进行人机交互的电子产品， 例如计算机、智能手机、普通手机、平板电脑等；所述网络设备包括但不限于计算机、网络主机、单个网络服务器、多个网络服务器集或多个服务器构成的云。
以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的各个模块可以是或者也可以不是物理上分开的。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性的劳动的情况下，即可以理解并实施。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到各实施方式可借助加必需的通用硬件平台的方式来实现，当然也可以通过硬件和软件结合的方式来实现。基于这样的理解，上述技术方案本质上或者说对现有技术做出贡献的部分可以以计算机产品的形式体现出来，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
最后应说明的是：以上实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。

Claims (10)

  1. 一种记录决策行为的方法，其中，包括：
    构建决策记录模型簇，所述决策记录模型簇包括不同的决策记录模型；
    所述不同的决策记录模型对决策事件数据集进行独立学习，在设定时间内共同生长；
    对所述决策记录模型簇考核，根据设定的淘汰规则对决策记录模型做管理。
  2. 根据权利要求1所述的方法，其中，所述决策事件数据集包括决策事件，所述决策事件包括依赖数据维度、操作数据维度和评价数据维度；同一决策记录模型簇的评价数据维度相同。
  3. 根据权利要求2所述的方法，其中，所述依赖数据维度包括决策事件所依赖的当前事实，事务状况的数据维度，记为X；所述操作数据维度包括决策事件的决策内容、决策后行为的数据维度，记为Y；所述评价数据维度包括决策事件的决策结果优劣评价指标、方式，记为Z；所述X，Y，Z构成决策记录模型。
  4. 根据权利要求1所述的方法，其中，所述考核包括在决策记录模型簇的各自独立的决策记录模型里输入相同决策事件各自的X值，输出各自历史最优或计算出的Y值和相应的Z值，并建立相应的淘汰规则。
  5. 根据权利要求4所述的方法，其中，所述淘汰规则包括但不限于：
    依据一段时间内的Z值平均值进行排序，对决策记录模型末位淘汰；
    对具有包含关系的决策记录模型之间，通过吸收或取代方式淘汰劣势决策记录模型；
    通过合并决策记录模型生成新的决策记录模型；
    通过模型塑身生成更简单的决策记录模型。
  6. 根据权利要求5所述的淘汰规则，其中：
    令决策记录模型a和决策记录模型b是决策记录模型簇中的两个决策记录模型,v＝a的Z的均值-b的Z的均值,
    当v>0且Z的方向性是越大越优时，称a优于b或b劣于a；
    当v>0且Z的方向性是越小越优时，称b优于a或a劣于b；
    当v>预设的显著值>0且Z的方向性是越大越优时，称a明显优于b或b明显劣于a；
    当v>预设的显著值>0且Z的方向性是越小越优时，称b明显优于a或a明显劣于b。
  7. 根据权利要求5所述的淘汰规则，其中，所述合并决策记录模型生成新的决策记录模型指：合并决策记录模型a和决策记录模型b生成决策记录模型c，{X} c＝{X} a∪{X} b，{Y} c＝{Y} a∪{Y} b，{Z} c＝{Z} a＝{Z} b。
  8. 根据权利要求5所述的淘汰规则，其中，决策记录模型塑身是指删除一个决策记录模型的{X}中的若干个数据维度，或删除{Y}中的若干个数据维度，生成新的更简单的决策记录模型。
  9. 一种记录决策行为的系统，其中，包括：
    模型建立模块，被配置为用于输入或建立决策记录模型，以形成决策记录模型簇；
    数据采集模块，被配置为用于采集或输入决策事件数据集，以供决策记录模型学习；
    模型分析模块，被配置为输入决策事件以考核决策记录模型簇，并淘汰劣势决策记录模型；
    预测模块，被配置为输入决策事件以输出操作决策和结果预测。
  10. 一种记录决策行为的设备，其中，包括：存储器、处理器；所述存储器上存储有可执行代码，当所述可执行代码被所述处理器执行时，使所述处理器执行如权利要求1至8中任一项所述的记录决策行为的方法。
PCT/CN2020/086253 2020-04-07 2020-04-23 一种记录决策行为的方法、系统和设备 WO2021203489A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010266520.2 2020-04-07
CN202010266520.2A CN113496335A (zh) 2020-04-07 2020-04-07 一种记录决策行为的方法、系统和设备

Publications (1)

Publication Number Publication Date
WO2021203489A1 true WO2021203489A1 (zh) 2021-10-14

Family

ID=77995650

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/086253 WO2021203489A1 (zh) 2020-04-07 2020-04-23 一种记录决策行为的方法、系统和设备

Country Status (2)

Country Link
CN (1) CN113496335A (zh)
WO (1) WO2021203489A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116542831B (zh) * 2023-07-07 2023-10-03 杭州海亮优教教育科技有限公司 招生数据的处理方法、装置、电子设备及存储介质

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105447519A (zh) * 2015-11-20 2016-03-30 携程计算机技术（上海）有限公司 基于特征选择的模型检测方法
US20160357807A1 (en) * 2015-06-02 2016-12-08 International Business Machines Corporation Horizontal Decision Tree Learning from Very High Rate Data Streams
CN107464025A (zh) * 2017-08-25 2017-12-12 智脑智能科技（苏州）有限公司 基于大数据的供应链需求预测与智能决策方法
CN107480686A (zh) * 2016-06-08 2017-12-15 阿里巴巴集团控股有限公司 一种筛选机器学习特征的方法和装置
CN107895501A (zh) * 2017-09-29 2018-04-10 大圣科技股份有限公司 基于海量驾驶视频数据训练的无人汽车驾驶决策方法
CN110070248A (zh) * 2018-01-19 2019-07-30 财团法人工业技术研究院 动态智能调度方法及装置
CN110399986A (zh) * 2019-06-24 2019-11-01 中水三立数据技术股份有限公司 一种泵站机组故障诊断系统的生成方法
CN110752942A (zh) * 2019-09-06 2020-02-04 平安科技（深圳）有限公司 告警信息的决策方法、装置、计算机设备及存储介质

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH1173323A (ja) * 1997-08-29 1999-03-16 Nippon Telegr & Teleph Corp <Ntt> 決定木生成方法及び決定木生成プログラムを記録した記録媒体
US20100223211A1 (en) * 2000-10-11 2010-09-02 Johnson Gregory A Decision service method and system
US20050096950A1 (en) * 2003-10-29 2005-05-05 Caplan Scott M. Method and apparatus for creating and evaluating strategies
US20090187425A1 (en) * 2007-09-17 2009-07-23 Arthur Solomon Thompson PDA software robots leveraging past history in seconds with software robots
JP6352165B2 (ja) * 2014-04-04 2018-07-04 パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカＰａｎａｓｏｎｉｃ Ｉｎｔｅｌｌｅｃｔｕａｌ Ｐｒｏｐｅｒｔｙ Ｃｏｒｐｏｒａｔｉｏｎ ｏｆ Ａｍｅｒｉｃａ 評価結果表示方法、評価結果表示装置及び評価結果表示プログラム
CN110047006A (zh) * 2019-04-22 2019-07-23 陈硕坚 基于神经网络和机器学习进行证券投资决策的方法及装置
CN109991987B (zh) * 2019-04-29 2023-08-04 北京智行者科技股份有限公司 自动驾驶决策方法及装置
CN110210718A (zh) * 2019-05-09 2019-09-06 厦门邑通软件科技有限公司 一种基于多维决策树群的提升产品合格率的方法

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160357807A1 (en) * 2015-06-02 2016-12-08 International Business Machines Corporation Horizontal Decision Tree Learning from Very High Rate Data Streams
CN105447519A (zh) * 2015-11-20 2016-03-30 携程计算机技术（上海）有限公司 基于特征选择的模型检测方法
CN107480686A (zh) * 2016-06-08 2017-12-15 阿里巴巴集团控股有限公司 一种筛选机器学习特征的方法和装置
CN107464025A (zh) * 2017-08-25 2017-12-12 智脑智能科技（苏州）有限公司 基于大数据的供应链需求预测与智能决策方法
CN107895501A (zh) * 2017-09-29 2018-04-10 大圣科技股份有限公司 基于海量驾驶视频数据训练的无人汽车驾驶决策方法
CN110070248A (zh) * 2018-01-19 2019-07-30 财团法人工业技术研究院 动态智能调度方法及装置
CN110399986A (zh) * 2019-06-24 2019-11-01 中水三立数据技术股份有限公司 一种泵站机组故障诊断系统的生成方法
CN110752942A (zh) * 2019-09-06 2020-02-04 平安科技（深圳）有限公司 告警信息的决策方法、装置、计算机设备及存储介质

Also Published As

Publication number Publication date
CN113496335A (zh) 2021-10-12

Similar Documents

Publication Publication Date Title
He et al. A counterfactual scenario simulation approach for assessing the impact of farmland preservation policies on urban sprawl and food security in a major grain-producing area of China
Clarke et al. The dynamics of urban spatial structure: the progress of a research programme
CN111461516B (zh) 一种智慧城市综合网格化管理系统平台
CN112418699A (zh) 资源分配方法、装置、设备及存储介质
Landau et al. Migration and population trends in Gauteng province 1996–2055
CN116384803A (zh) 一种城市数据分析方法、系统、电子设备和存储介质
Alnsour The effectiveness of urban management in Jordanian municipalities
Krapp et al. Housing policies in the European Union
WO2021203489A1 (zh) 一种记录决策行为的方法、系统和设备
Zhang et al. A robust optimization model for green regional logistics network design with uncertainty in future logistics demand
Georgescu et al. The race for making up the list of emergent smart cities. An Eastern European country’s approach
Bulut et al. Decentralization and poverty reduction: Opportunities and challenges in Kenya
Schletz et al. Nested climate accounting for our atmospheric commons—digital technologies for trusted interoperability across fragmented systems
Jannah et al. Designing the Smart Health Function towards Puskesmas (Citizen Health Centre) based on Smart City Concept
Xu et al. A study on the risk assessment of water conservancy scenic spot PPP projects
Chutimaskul E-government analysis and modeling
Marasović et al. Implementing a WebGIS Solution in the City of Sinj in Accordance with the Needs of the City Stakeholders
Frimpong Managing plan implementation in the Asante Akyem South District Assembly: Capacity issues and challenges
Marasović et al. Uspostava webGIS rješenja u gradu Sinju u skladu s potrebama gradskih dionika
Raymer et al. A multiregional sources of growth model for school enrolment projections
Munzhelele Knowledge management and service delivery: A knowledge management model for the housing sector
Mohammadi The Pathology of Urban Master Plans in Iran
Nuwagaba et al. Public Private Partnership Projects and Regional Cooperation in The Horn of Africa and the East African Region in the era of COVID-19 pandemic
Sukmadewi et al. Public Service Agency: Business-Based Innovation in Public Institutions:(Study of Design of Government Institutions in The Field of Communications And Informatics)
Pazos-Vidal Scotland: a thriving environment for policy entrepreneurship?

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: 20930591

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: 20930591

Country of ref document: EP

Kind code of ref document: A1