WO2023142351A1 - 权重的调整方法和装置、存储介质及电子装置 - Google Patents

权重的调整方法和装置、存储介质及电子装置 Download PDF

Info

Publication number
WO2023142351A1
WO2023142351A1 PCT/CN2022/100559 CN2022100559W WO2023142351A1 WO 2023142351 A1 WO2023142351 A1 WO 2023142351A1 CN 2022100559 W CN2022100559 W CN 2022100559W WO 2023142351 A1 WO2023142351 A1 WO 2023142351A1
Authority
WO
WIPO (PCT)
Prior art keywords
continuous
action
actions
weight
continuous action
Prior art date
Application number
PCT/CN2022/100559
Other languages
English (en)
French (fr)
Inventor
刘建国
周杰
Original Assignee
青岛海尔科技有限公司
海尔智家股份有限公司
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 青岛海尔科技有限公司, 海尔智家股份有限公司 filed Critical 青岛海尔科技有限公司
Publication of WO2023142351A1 publication Critical patent/WO2023142351A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning

Definitions

  • the present disclosure relates to the communication field, and in particular, to a weight adjustment method and device, a storage medium, and an electronic device.
  • Zhijia Brain is a ubiquitous, safe and reliable system that naturally communicates with users, actively and intimately serves users, and has the ability of continuous learning and evolution. Prediction of the user's behavior habits is an important function among the various capabilities of the brain. Before making predictions, the brain first needs to learn the user's behavior habits and generate the user's behavior habits. If we judge through the brain that the user is likely to operate other behaviors when the user is operating the device, we will actively prompt the user and perform operations for the user, which will bring a great user experience.
  • the behavior habit mining of the existing technology is mainly to analyze the user's offline behavior habits, that is, analyze the historical data before yesterday, perform some preference processing through the historical data, and obtain long-term behavior by means of behavior statistics.
  • the method of analyzing user behavior through offline behavior lacks the behavior change offset of the user on the day, which will lead to inaccurate prediction of user behavior habits.
  • Embodiments of the present disclosure provide a weight adjustment method and device, a storage medium, and an electronic device, so as to at least solve the problem in the related art of predicting user behavior habits based on historical data, resulting in inaccurate prediction results.
  • a weight adjustment method including: obtaining the first continuous action of the target object within a preset time period, wherein the preset time period is a time period before the current moment Obtaining the behavior data set of the target object, wherein the behavior data set includes: a plurality of second continuous actions, and a plurality of first weights respectively corresponding to the plurality of second continuous actions; the second continuous actions and The first weights are in one-to-one correspondence; the plurality of first weights respectively corresponding to the plurality of second continuous actions are adjusted according to the first continuous action.
  • a weight adjustment device including: a first acquisition module, configured to acquire the first continuous action of the target object within a preset time period, wherein the preset The time period is the time period before the current moment; the second acquisition module is configured to acquire the behavior data set of the target object, wherein the behavior data set includes: a plurality of second continuous actions, and a plurality of second continuous actions A plurality of corresponding first weights; the second continuous actions correspond to the first weights one by one; an adjustment module is configured to adjust the corresponding multiples of the plurality of second continuous actions according to the first continuous actions a first weight.
  • a computer-readable storage medium where a computer program is stored in the computer-readable storage medium, wherein the computer program is configured to perform the above-mentioned weight adjustment when running method.
  • an electronic device including a memory, a processor, and a computer program stored on the memory and operable on the processor, wherein the above-mentioned processor executes the above-mentioned How to adjust the weight.
  • the first continuous action of the target object within a preset time period is obtained, wherein the preset time period is a time period before the current moment;
  • the behavior data set of the target object is obtained, wherein the The behavior data set includes: a plurality of second continuous actions, and a plurality of first weights respectively corresponding to the plurality of second continuous actions; the second continuous actions are in one-to-one correspondence with the first weights; according to the first The continuous action adjusts the plurality of first weights respectively corresponding to the plurality of second continuous actions; the above technical solution solves the problem of predicting user behavior habits based on historical data, resulting in inaccurate prediction results.
  • Combining the real-time learned behavior and offline behavior to get the final user behavior habits will make behavior prediction more reliable and accurate.
  • FIG. 1 is a block diagram of the hardware structure of a computer terminal according to a weight adjustment method according to an embodiment of the present disclosure
  • FIG. 2 is a flowchart of a method for adjusting weights according to an embodiment of the present disclosure
  • FIG. 3 is an overall block diagram of a method for adjusting weights according to an embodiment of the present disclosure
  • FIG. 4 is an overall flowchart of a weight adjustment method according to an embodiment of the present disclosure
  • FIG. 5 is a detailed diagram of a weight adjustment method according to an embodiment of the present disclosure.
  • FIG. 6 is a graph showing changes in curves of weights of continuous behaviors according to an embodiment of the present disclosure
  • Fig. 7 is a structural block diagram of an apparatus for adjusting weights according to an embodiment of the present disclosure.
  • FIG. 1 is a block diagram of a hardware structure of a computer terminal according to a method for adjusting weights according to an embodiment of the present disclosure.
  • the computer terminal may include one or more (only one is shown in Figure 1) processors 102 (processors 102 may include but not limited to processing devices such as microprocessor MCU or programmable logic device FPGA, etc.) and a memory 104 for storing data.
  • processors 102 may include but not limited to processing devices such as microprocessor MCU or programmable logic device FPGA, etc.
  • the above-mentioned computer terminal may further include a transmission device 106 and an input and output device 108 for communication functions.
  • FIG. 1 is only for illustration, and it does not limit the structure of the above computer terminal.
  • the computer terminal may also include more or less components than those shown in FIG. 1 , or have a different configuration with functions equivalent to those shown in FIG. 1 or more functions than those shown in FIG. 1 .
  • the memory 104 can be set to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the weight adjustment method in the embodiment of the present disclosure, and the processor 102 runs the computer program stored in the memory 104, thereby Executing various functional applications and data processing is to realize the above-mentioned method.
  • the memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory.
  • the memory 104 may further include a memory that is remotely located relative to the processor 102, and these remote memories may be connected to a computer terminal through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
  • Transmission device 106 is configured to receive or transmit data via a network.
  • the specific example of the above-mentioned network may include a wireless network provided by the communication provider of the computer terminal.
  • the transmission device 106 includes a network interface controller (NIC for short), which can be connected to other network devices through a base station so as to communicate with the Internet.
  • the transmission device 106 may be a radio frequency (Radio Frequency, referred to as RF) module, which is configured to communicate with the Internet in a wireless manner.
  • RF radio frequency
  • FIG. 2 is a flowchart of a weight adjustment method according to an embodiment of the present disclosure. The process includes the following steps:
  • Step S202 acquiring the first continuous action of the target object within a preset time period, wherein the preset time period is a time period before the current moment;
  • Step S204 acquiring the behavior data set of the target object, wherein the behavior data set includes: a plurality of second continuous actions, and a plurality of first weights respectively corresponding to the plurality of second continuous actions; Actions are in one-to-one correspondence with the first weight;
  • Step S206 adjusting a plurality of first weights respectively corresponding to the plurality of second continuous actions according to the first continuous action.
  • the first continuous action of the target object within the preset time period is obtained, wherein the preset time period is a time period before the current moment; the behavior data set of the target object is obtained, wherein the behavior data
  • the set includes: a plurality of second continuous actions, and a plurality of first weights respectively corresponding to the plurality of second continuous actions; the second continuous actions are in one-to-one correspondence with the first weights; adjusted according to the first continuous actions
  • the plurality of first weights respectively corresponding to the plurality of second continuous actions solves the problem in the related art of predicting user behavior habits based on historical data, resulting in inaccurate prediction results.
  • the combination of user behavior and offline behavior can get the final user behavior habits, which will make the behavior prediction more reliable and accurate.
  • adjusting a plurality of first weights respectively corresponding to the plurality of second continuous actions according to the first continuous action includes: determining whether there is a weight corresponding to the plurality of second continuous actions A third continuous action that is all and/or partially consistent with the first continuous action; adjust the multiple second continuous actions according to whether there is a third continuous action that is fully and/or partially consistent with the first continuous action. A plurality of first weights respectively corresponding to the second continuous actions.
  • the consistency of the continuous actions adjusts the plurality of first weights respectively corresponding to the historical continuous actions.
  • the respective corresponding values of the plurality of second continuous actions are adjusted.
  • a plurality of first weights including: in the case that there is a third continuous action that is all consistent with the first continuous action in the plurality of second continuous actions, increasing the first weight of the third continuous action to second weight; if there is a third continuous action partially consistent with the first continuous action among the plurality of second continuous actions, increasing the first weight of the third continuous action to the second weight, and adding the first continuous action to the behavior data set, and assigning a default weight to the first continuous action; among the plurality of second continuous actions, there are In the case of multiple third continuous actions, the first weights respectively corresponding to the multiple third continuous actions are respectively increased to the second weights.
  • the multiple second continuous actions include: action 1): action A, action B, action C; action 2): action A, action B; action 3): action A, action C, in the first
  • action 1) action A, action B, action C, and action D
  • add the weight of action 1) and action 2) to the second weight and add in the behavior data set: action 4): action A, action B , action C, and action D, and give default weight to action 4).
  • the weight of action 2) is increased to the second weight.
  • a plurality of first weights comprising: adding the first continuous action to the behavior if there is no third continuous action in the plurality of second continuous actions that all agree with the first continuous action data set, and assign a default weight to the first continuous action; if there is no third continuous action partially consistent with the first continuous action among the plurality of second continuous actions, the multiple a plurality of first weights respectively corresponding to the second continuous actions; in the case that there is no third continuous action that is wholly or partially consistent with the first continuous action in the plurality of second continuous actions, the first continuous action is set to A continuous action is added to the behavior data set, a default weight is assigned to the first continuous action, and multiple first weights respectively corresponding to the multiple second continuous actions are not adjusted.
  • the multiple second continuous actions include: action 1): action A, action B, action C; action 2): action A, action B; action 3): action A, action C, in the first
  • the weights of action 1), action 2), and action 3) are not adjusted, and the weights of action 4) are added in the behavior data set: action A, action D, and action 4) Assign default weights.
  • the behavior data set After adjusting a plurality of first weights respectively corresponding to the plurality of second continuous actions according to the first continuous action, it is determined whether the behavior data set has reached a decay period; When the set reaches the decay period, obtain the current moment and the last decay moment; determine the decay factor of the behavior dataset according to the current moment and the last decay moment; adjust the behavior dataset according to the decay factor The first weight of the plurality of second continuous actions of .
  • the attenuation factor of the behavior data set is determined according to the following formula: Wherein, t is the current moment, t0 is the last attenuation moment, and ⁇ is a variable parameter; the attenuation factor is multiplied by the adjusted first weight to obtain the third of multiple second continuous actions in the behavior data set. Weights.
  • the embodiment of the present disclosure will periodically refresh the behavior, and if the behavior reaches the decay period, the decay behavior will be unified.
  • the decay factor is designed as: Where t is the current time in minutes, t0 is the previous decay time, and ⁇ is a variable parameter. This formula satisfies the characteristics that the decay is faster when the distance is closer, and the decay is slower when the distance is farther away. , in line with the memory of the human brain.
  • the first action of the target object is acquired, wherein the first An action is an individual action used to indicate the target object; a fourth continuous action including the first action and a fourth weight corresponding to the fourth continuous action are determined in the target data set; according to the fourth continuous The action and the fourth weight predict an action to be performed by the target object.
  • the user judges that the user is likely to operate other behaviors through the above data weight when operating the device, and proactively prompting the user and performing operations for the user will bring a great user experience.
  • the real-time learning behavior and offline behavior are combined to obtain the final user behavior habits, which will make behavior prediction more reliable and accurate.
  • FIG. 3 is an overall block diagram of a weight adjustment method according to an embodiment of the present disclosure
  • FIG. 4 is an overall flowchart of a weight adjustment method according to an embodiment of the present disclosure, as As shown in Figure 3 and Figure 4, the specific steps are as follows:
  • Step S301 initialization of historical long-term behavior
  • Step S302 real-time behavior analysis
  • behavior A at time 1 in Figure 5 continuous behaviors of A and B at time 2; A, B, B at time 3 C continuous behavior;
  • Step S303 correct long-term behavior
  • the behavior weight will increase, as shown in Figure 5, the weight of B and C continuous behavior at time 3 increases; the increase of A and B continuous behavior at time 2 2) Refresh the behavior regularly, and if the behavior reaches the decay period, the decay behavior will be unified.
  • the decay factor is designed as Where t is the current time in minutes, t0 is the previous decay time, and ⁇ is a variable parameter. This formula satisfies the characteristics that the decay is faster when the distance is closer, and the decay is slower when the distance is farther away. , consistent with the memory of the human brain.
  • Step S304 new behavior record.
  • the new behavior is updated into a new behavior record.
  • FIG. 6 is a curve change graph of the weight of the continuous behavior according to an embodiment of the present disclosure. Under different attenuation factors, the weight of the continuous behavior has different change curves.
  • the behavior habit mining of the existing technology is mainly to analyze the user's offline behavior habits, that is, analyze the historical data before yesterday, perform some preference processing through the historical data, and obtain long-term behavior by means of behavior statistics.
  • the method of analyzing user behavior through offline behavior lacks the behavior change offset of the user on the day, which will lead to inaccurate prediction of user behavior habits. Therefore, in the embodiments of the present disclosure, the real-time learned behavior and offline behavior are combined to obtain the final user behavior habits, which will make the behavior prediction more reliable and accurate.
  • the method according to the above embodiments can be implemented by means of software plus a necessary general-purpose hardware platform, and of course also by hardware, but in many cases the former is better implementation.
  • the technical solution of the present disclosure can be embodied in the form of a software product in essence or the part that contributes to the prior art, and the computer software product is stored in a storage medium (such as ROM/RAM, disk, CD) contains several instructions to make a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) execute the method of each embodiment of the present disclosure.
  • a storage medium such as ROM/RAM, disk, CD
  • a weight adjustment device is also provided, and the weight adjustment device is configured to implement the above embodiments and preferred implementation manners, and what has already been described will not be repeated.
  • the term "module” may be a combination of software and/or hardware that realizes a predetermined function.
  • the devices described in the following embodiments are preferably implemented in software, implementations in hardware, or a combination of software and hardware are also possible and contemplated.
  • Fig. 7 is a structural block diagram of a weight adjustment device according to an embodiment of the present disclosure; as shown in Fig. 7 , it includes:
  • the first acquisition module 72 is configured to acquire the first continuous action of the target object within a preset time period, wherein the preset time period is a time period before the current moment;
  • the second acquisition module 74 is configured to acquire the behavior data set of the target object, wherein the behavior data set includes: a plurality of second continuous actions, and a plurality of first weights respectively corresponding to the plurality of second continuous actions;
  • the second continuous action is in one-to-one correspondence with the first weight;
  • the adjustment module 76 is configured to adjust a plurality of first weights respectively corresponding to the plurality of second continuous actions according to the first continuous action.
  • the first continuous action of the target object within a preset time period is obtained, wherein the preset time period is a time period before the current moment; the behavior data set of the target object is obtained, wherein the behavior data
  • the set includes: a plurality of second continuous actions, and a plurality of first weights respectively corresponding to the plurality of second continuous actions; the second continuous actions are in one-to-one correspondence with the first weights; adjusted according to the first continuous actions
  • the plurality of first weights respectively corresponding to the plurality of second continuous actions solves the problem in the related art of predicting user behavior habits based on historical data, resulting in inaccurate prediction results.
  • the combination of user behavior and offline behavior can get the final user behavior habits, which will make the behavior prediction more reliable and accurate.
  • the adjustment module 76 is configured to determine whether there is a third continuous action that is wholly and/or partially consistent with the first continuous action among the plurality of second continuous actions; Whether there is a third continuous action that is completely and/or partially consistent with the first continuous action in the second continuous action adjusts the plurality of first weights respectively corresponding to the plurality of second continuous actions.
  • the adjustment module 76 is configured to, if there is a third continuous action that is all consistent with the first continuous action among the plurality of second continuous actions, set the third continuous action to The first weight of the second continuous action is increased to the second weight; in the case that there is a third continuous action partially consistent with the first continuous action in the plurality of second continuous actions, the first weight of the third continuous action adding to the second weight, and adding the first continuous action to the behavior data set, and assigning a default weight to the first continuous action; among the plurality of second continuous actions, there are In the case of a plurality of third continuous actions in which all or part of the continuous actions are consistent, the first weights respectively corresponding to the plurality of third continuous actions are respectively increased to the second weights.
  • the adjustment module 76 is configured to set the first continuous action to An action is added to the behavior data set, and a default weight is assigned to the first continuous action; if there is no third continuous action partially consistent with the first continuous action among the plurality of second continuous actions , do not adjust the plurality of first weights respectively corresponding to the plurality of second continuous actions; there is no third continuous action that is wholly or partially consistent with the first continuous action in the plurality of second continuous actions Next, adding the first continuous action to the behavior data set, assigning a default weight to the first continuous action, and not adjusting the plurality of first weights respectively corresponding to the plurality of second continuous actions.
  • the adjustment module 76 is configured to determine whether the behavior data set has reached the decay period; if the behavior data set has reached the decay period, obtain the current moment and the last decay moment; according to the The current moment and the last decay moment determine the decay factor of the behavior data set; and adjust the first weights of multiple second continuous actions in the behavior data set according to the decay factor.
  • the adjustment module 76 is configured to determine the attenuation factor of the behavior data set according to the following formula: Wherein, t is the current moment, t0 is the last attenuation moment, and ⁇ is a variable parameter; the attenuation factor is multiplied by the adjusted first weight to obtain the third of multiple second continuous actions in the behavior data set. Weights.
  • the first obtaining module is configured to obtain a first action of the target object, wherein the first action is an individual action used to indicate the target object; in the target data set Determining a fourth continuous action including the first action and a fourth weight corresponding to the fourth continuous action; predicting an action to be performed by the target object according to the fourth continuous action and the fourth weight.
  • An embodiment of the present disclosure also provides a storage medium, the storage medium includes a stored program, wherein the above-mentioned program executes any one of the above-mentioned methods when running.
  • the above-mentioned storage medium may be configured to store program codes for performing the following steps:
  • the behavior data set includes: a plurality of second continuous actions, and a plurality of first weights respectively corresponding to the plurality of second continuous actions; the second continuous actions one-to-one correspondence with the first weight;
  • Embodiments of the present disclosure also provide an electronic device, including a memory and a processor, where a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.
  • the above-mentioned electronic device may further include a transmission device and an input-output device, wherein the transmission device is connected to the above-mentioned processor, and the input-output device is connected to the above-mentioned processor.
  • the above-mentioned processor may be configured to execute the following steps through a computer program:
  • the behavior data set includes: a plurality of second continuous actions, and a plurality of first weights respectively corresponding to the plurality of second continuous actions; the second continuous actions one-to-one correspondence with the first weight;
  • the above-mentioned storage medium may include but not limited to: U disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), Various media that can store program codes such as removable hard disks, magnetic disks, or optical disks.
  • ROM read-only memory
  • RAM random access memory
  • Various media that can store program codes such as removable hard disks, magnetic disks, or optical disks.
  • each module or each step of the above-mentioned disclosure can be realized by a general-purpose computing device, and they can be concentrated on a single computing device, or distributed in a network composed of multiple computing devices Alternatively, they may be implemented in program code executable by a computing device so that they may be stored in a storage device to be executed by a computing device, and in some cases in an order different from that shown here
  • the steps shown or described are carried out, or they are separately fabricated into individual integrated circuit modules, or multiple modules or steps among them are fabricated into a single integrated circuit module for implementation.
  • the present disclosure is not limited to any specific combination of hardware and software.

Abstract

本公开提供了一种权重的调整方法和装置、存储介质及电子装置，其中，上述方法包括：获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重。

Description

权重的调整方法和装置、存储介质及电子装置
本公开要求于2022年01月28日提交中国专利局、申请号为202210109432.0、发明名称“权重的调整方法和装置、存储介质及电子装置”的中国专利申请的优先权，其全部内容通过引用结合在本公开中。
技术领域
本公开涉及通信领域，具体而言，涉及一种权重的调整方法和装置、存储介质及电子装置。
背景技术
智家大脑是一个无处不在，与用户自然交流，为用户主动贴心服务，安全可靠的系统，并且拥有持续学习进化的能力。在大脑的各项能力中用户的行为习惯预测是一个重要的功能，在进行预测前首先大脑需要学习用户的行为习惯并生成用户的行为习惯。如果用户在操作设备的时候我们通过大脑判断出用户还有很大可能操作其他行为，我们主动提示用户并给用户进行操作将带来极大的用户体验。
现有技术的行为习惯挖掘主要是分析用户的离线行为习惯，即在今天分析昨天以前的历史数据，通过历史数据进行一些偏好处理，行为统计的方式得到长期行为。这样通过离线行为分析用户行为的方法就缺少了用户当天的行为变化偏移，会导致用户行为习惯的预测不准确。
针对相关技术中，根据历史数据对用户行为习惯进行预测，导致预测结果不准确的问题，尚未提出有效的解决方案。
发明内容
本公开实施例提供了一种权重的调整方法和装置、存储介质及电子装置，以至少解决相关技术中，根据历史数据对用户行为习惯进行预测，导致预测结果不准确的问题。
根据本公开实施例的一个实施例，提供了一种权重的调整方法，包括：获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重。
根据本公开实施例的另一个实施例，还提供了一种权重的调整装置，包括：第一获取模块，设置为获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；第二获取模块，设置为获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；调整模块，设置为根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重。
根据本公开实施例的又一方面，还提供了一种计算机可读的存储介质，该计算机可读的存储介质中存储有计算机程序，其中，该计算机程序被设置为运行时执行上述权重的调整方法。
根据本公开实施例的又一方面，还提供了一种电子装置，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，其中，上述处理器通过计算机程序执行上述的权重的调整方法。
在本公开实施例中，获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重；采用上述技术方案，解决了根据历史数据对用户行为习惯进行预测，导致预测结果不准确的问题，本公开实施例中将实时学习的行为和离线行为相结合得到最终的用户行为习惯，这样会使行为预测更加可靠准确。
附图说明
此处所说明的附图用来提供对本公开的进一步理解，构成本公开的一部分，本公开的示意性实施例及其说明用于解释本公开，并不构成对本公开的不当限定。在附图中：
图1是本公开实施例的一种权重的调整方法的计算机终端的硬件结构框图；
图2是根据本公开实施例的权重的调整方法的流程图；
图3是根据本公开实施例的权重的调整方法的整体框图；
图4是根据本公开实施例的权重的调整方法的整体流程图；
图5是根据本公开实施例的权重的调整方法的明细图；
图6是根据本公开实施例的连续行为的权重的曲线变化图；
图7是根据本公开实施例的一种权重的调整装置的结构框图。
具体实施方式
为了使本技术领域的人员更好地理解本公开方案，下面将结合本公开实施例中的附图，对本公开实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本公开一部分的实施例，而不是全部的实施例。基于本公开中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都应当属于本公开保护的范围。
需要说明的是，本公开的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便这里描述的本公开的实施例能够以除了在这里图示或描述的那些以外的顺序实施。此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排 他的包含，例如，包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。
本公开实施例所提供的方法实施例可以在移动终端、计算机终端或者类似的运算装置中执行。以运行在计算机终端上为例，图1是本公开实施例的一种权重的调整方法的计算机终端的硬件结构框图。如图1所示，计算机终端可以包括一个或多个(图1中仅示出一个)处理器102(处理器102可以包括但不限于微处理器MCU或可编程逻辑器件FPGA等的处理装置)和用于存储数据的存储器104，在一个示例性实施例中，上述计算机终端还可以包括用于通信功能的传输设备106以及输入输出设备108。本领域普通技术人员可以理解，图1所示的结构仅为示意，其并不对上述计算机终端的结构造成限定。例如，计算机终端还可包括比图1中所示更多或者更少的组件，或者具有与图1所示等同功能或比图1所示功能更多的不同的配置。
存储器104可设置为存储计算机程序，例如，应用软件的软件程序以及模块，如本公开实施例中的权重的调整方法对应的计算机程序，处理器102通过运行存储在存储器104内的计算机程序，从而执行各种功能应用以及数据处理，即实现上述的方法。存储器104可包括高速随机存储器，还可包括非易失性存储器，如一个或者多个磁性存储装置、闪存、或者其他非易失性固态存储器。在一些实例中，存储器104可进一步包括相对于处理器102远程设置的存储器，这些远程存储器可以通过网络连接至计算机终端。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
传输设备106设置为经由一个网络接收或者发送数据。上述的网络具体实例可包括计算机终端的通信供应商提供的无线网络。在一个实例中，传输设备106包括一个网络适配器(Network Interface Controller，简称为NIC)，其可通过基站与其他网络设备相连从而可与互联网进行通讯。在一 个实例中，传输设备106可以为射频(Radio Frequency，简称为RF)模块，其设置为通过无线方式与互联网进行通讯。
在本实施例中提供了一种权重的调整方法，应用于上述计算机终端，图2是根据本公开实施例的权重的调整方法的流程图，该流程包括如下步骤：
步骤S202，获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；
步骤S204，获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；
步骤S206，根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重。
通过上述步骤，获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重，解决了相关技术中，根据历史数据对用户行为习惯进行预测，导致预测结果不准确的问题，进而本公开实施例中将实时学习的行为和离线行为相结合得到最终的用户行为习惯，这样会使行为预测更加可靠准确。
在一个示例性实施例中，根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重，包括：确定所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作；根据所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作调整所述多个第二连续动作分别对应的多个第一权重。
也就是说，根据当前的连续动作(相当于上述实施例中的第一连续动 作)和历史的连续动作(相当于上述实施例中的第二连续动作)进行对比，根据当前的连续动作和历史的连续动作的一致性调整所述历史的连续动作分别对应的多个第一权重。
在一个示例性实施例中，根据所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作调整所述多个第二连续动作分别对应的多个第一权重，包括：在所述多个第二连续动作中存在与所述第一连续动作全部一致的第三连续动作的情况下，将所述第三连续动作的第一权重增加至第二权重；在所述多个第二连续动作中存在与所述第一连续动作部分一致的第三连续动作的情况下，将所述第三连续动作的第一权重增加至第二权重，以及将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重；在所述多个第二连续动作中存在与所述第一连续动作全部和部分一致的多个第三连续动作的情况下，分别将所述多个第三连续动作分别对应的第一权重增加至第二权重。
举例来讲，在多个第二连续动作中包括：动作1)：动作A、动作B、动作C；动作2)：动作A、动作B；动作3)：动作A、动作C，在第一连续动作为动作A、动作B、动作C、动作D的情况下，将动作1)；动作2)的权重增加至第二权重，并且在行为数据集中增加：动作4)：动作A、动作B、动作C、动作D，并为动作4)赋予默认权重。在第一连续动作为动作A、动作B的情况下，将动作2)的权重增加至第二权重。需要说明的是，上述实施例仅是为了更好的理解本公开，本公开实施例对动作顺序不做限定。
在一个示例性实施例中，根据所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作调整所述多个第二连续动作分别对应的多个第一权重，包括：在所述多个第二连续动作中不存在与所述第一连续动作全部一致的第三连续动作的情况下，将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重；在所述多个第二连续动作中不存在与所述第一连续动作部分一致的第三连 续动作的情况下，不调整所述多个第二连续动作分别对应的多个第一权重；在所述多个第二连续动作中不存在与所述第一连续动作全部和部分一致的第三连续动作的情况下，将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重，以及不调整所述多个第二连续动作分别对应的多个第一权重。
举例来讲，在多个第二连续动作中包括：动作1)：动作A、动作B、动作C；动作2)：动作A、动作B；动作3)：动作A、动作C，在第一连续动作为动作A、动作D的情况下，不调整动作1)；动作2)；动作3)的权重，并且在行为数据集中增加：动作4)：动作A、动作D，并为动作4)赋予默认权重。需要说明的是，上述实施例仅是为了更好的理解本公开，本公开实施例对动作顺序不做限定。
在一个示例性实施例中，根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重之后，确定所述行为数据集是否到达衰减周期；在所述行为数据集到达衰减周期的情况下，获取当前时刻、上一次衰减时刻；根据所述当前时刻、所述上一次衰减时刻确定所述行为数据集的衰减因子；根据所述衰减因子调整所述行为数据集中的多个第二连续动作的第一权重。
具体的，根据以下公式确定所述行为数据集的衰减因子：
Figure PCTCN2022100559-appb-000001
其中，t为当前时刻、t0为上一次衰减时刻，α为可变参数；将所述衰减因子乘调整后的第一权重，以得到所述行为数据集中的多个第二连续动作的第三权重。
也就是说，本公开实施例会定期刷新行为，如果行为到了衰减周期则统一衰减行为，为了让长期行为和短期行为对接数据平滑，衰减因子设计为：
Figure PCTCN2022100559-appb-000002
其中t为当前时刻时间，单位为分钟，t0为上一衰减时刻时间，α为可变参数，此公式满足了在越近的情况下衰减越快，在越远 的情况下衰减越慢的特性，符合人脑的记忆。
在一个示例性实施例中，根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重之后，获取所述目标对象的第一动作，其中，所述第一动作为用于指示所述目标对象的单独动作；在所述目标数据集中确定包含所述第一动作的第四连续动作和所述第四连续动作对应的第四权重；根据所述第四连续动作和所述第四权重预测所述目标对象的待执行动作。
也就是说，根据调整后的权重，用户在操作设备的时候通过上述数据权重判断出用户还有很大可能操作其他行为，主动提示用户并给用户进行操作将带来极大的用户体验。本方案中将实时学习的行为和离线行为相结合得到最终的用户行为习惯，这样会使行为预测更加可靠准确。
为了更好的理解上述权重的调整方法的过程，以下再结合可选实施例对上述权重的调整的实现方法流程进行说明，但不用于限定本公开实施例的技术方案。
在本实施例中提供了一种权重的调整方法，图3是根据本公开实施例的权重的调整方法的整体框图、图4是根据本公开实施例的权重的调整方法的整体流程图，如图3和图4所示，具体如下步骤：
步骤S301：历史长期行为初始化；
具体的，加载用户历史行为作为初始化行为记录，如图5中的长期行为初始化。
步骤S302：实时行为分析；
具体的，实时接入用户行为，并记录进入算法的每一个行为，并形成行为序列，如图5中时刻1中的A行为；时刻2的A、B连续行为；时刻3的A、B、C连续行为；
步骤S303：纠正长期行为；
具体的，1)处理前后连续行为，如果在历史行为中有相同行为则行为权重增大，如图5中的时刻3的B，C连续行为权重增大；时刻2的增加A、B连续行为的权重；2)定期刷新行为，如果行为到了衰减周期则统一衰减行为，为了让长期行为和短期行为对接数据平滑，衰减因子设计为
Figure PCTCN2022100559-appb-000003
其中t为当前时刻时间，单位为分钟，t0为上一衰减时刻时间，α为可变参数，此公式满足了在越近的情况下衰减越快，在越远的情况下衰减越慢的特性，符合人脑的记忆。
步骤S304：新行为记录。
具体的，新行为更新到新的行为记录中。
如图6所示，图6是根据本公开实施例的连续行为的权重的曲线变化图，不同的衰减因子下，连续行为的权重不同的变化曲线。
现有技术的行为习惯挖掘主要是分析用户的离线行为习惯，即在今天分析昨天以前的历史数据，通过历史数据进行一些偏好处理，行为统计的方式得到长期行为。这样通过离线行为分析用户行为的方法就缺少了用户当天的行为变化偏移，会导致用户行为习惯的预测不准确。因此本公开实施例中将实时学习的行为和离线行为相结合得到最终的用户行为习惯，这样会使行为预测更加可靠准确。
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根据上述实施例的方法可借助软件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，但很多情况下前者是更佳的实施方式。基于这样的理解，本公开的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软件产品的形式体现出来，该计算机软件产品存储在一个存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，或者网络设备等)执行本公开各个实施例的方法。
在本实施例中还提供了权重的调整装置，该权重的调整装置设置为实现上述实施例及优选实施方式，已经进行过说明的不再赘述。如以下所使用的，术语“模块”可以实现预定功能的软件和/或硬件的组合。尽管以下实施例所描述的装置较佳地以软件来实现，但是硬件，或者软件和硬件的组合的实现也是可能并被构想的。
图7是根据本公开实施例的一种权重的调整装置的结构框图；如图7所示，包括：
第一获取模块72，设置为获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；
第二获取模块74，设置为获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；
调整模块76，设置为根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重。
通过上述装置，获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重，解决了相关技术中，根据历史数据对用户行为习惯进行预测，导致预测结果不准确的问题，进而本公开实施例中将实时学习的行为和离线行为相结合得到最终的用户行为习惯，这样会使行为预测更加可靠准确。
在一个示例性实施例中，调整模块76，设置为确定所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作；根据所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作调整所述多个第二连续动作分别对应的多个第一 权重。
在一个示例性实施例中，调整模块76，设置为在所述多个第二连续动作中存在与所述第一连续动作全部一致的第三连续动作的情况下，将所述第三连续动作的第一权重增加至第二权重；在所述多个第二连续动作中存在与所述第一连续动作部分一致的第三连续动作的情况下，将所述第三连续动作的第一权重增加至第二权重，以及将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重；在所述多个第二连续动作中存在与所述第一连续动作全部和部分一致的多个第三连续动作的情况下，分别将所述多个第三连续动作分别对应的第一权重增加至第二权重。
在一个示例性实施例中，调整模块76，设置为在所述多个第二连续动作中不存在与所述第一连续动作全部一致的第三连续动作的情况下，将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重；在所述多个第二连续动作中不存在与所述第一连续动作部分一致的第三连续动作的情况下，不调整所述多个第二连续动作分别对应的多个第一权重；在所述多个第二连续动作中不存在与所述第一连续动作全部和部分一致的第三连续动作的情况下，将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重，以及不调整所述多个第二连续动作分别对应的多个第一权重。
在一个示例性实施例中，调整模块76，设置为确定所述行为数据集是否到达衰减周期；在所述行为数据集到达衰减周期的情况下，获取当前时刻、上一次衰减时刻；根据所述当前时刻、所述上一次衰减时刻确定所述行为数据集的衰减因子；根据所述衰减因子调整所述行为数据集中的多个第二连续动作的第一权重。
在一个示例性实施例中，调整模块76，设置为根据以下公式确定所述行为数据集的衰减因子：
Figure PCTCN2022100559-appb-000004
其中，t为当前时刻、t0为上一次 衰减时刻，α为可变参数；将所述衰减因子乘调整后的第一权重，以得到所述行为数据集中的多个第二连续动作的第三权重。
在一个示例性实施例中，第一获取模块，设置为获取所述目标对象的第一动作，其中，所述第一动作为用于指示所述目标对象的单独动作；在所述目标数据集中确定包含所述第一动作的第四连续动作和所述第四连续动作对应的第四权重；根据所述第四连续动作和所述第四权重预测所述目标对象的待执行动作。
本公开的实施例还提供了一种存储介质，该存储介质包括存储的程序，其中，上述程序运行时执行上述任一项的方法。
可选地，在本实施例中，上述存储介质可以被设置为存储用于执行以下步骤的程序代码：
S1，获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；
S2，获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；
S3，根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重。
本公开的实施例还提供了一种电子装置，包括存储器和处理器，该存储器中存储有计算机程序，该处理器被设置为运行计算机程序以执行上述任一项方法实施例中的步骤。
可选地，上述电子装置还可以包括传输设备以及输入输出设备，其中，该传输设备和上述处理器连接，该输入输出设备和上述处理器连接。
可选地，在本实施例中，上述处理器可以被设置为通过计算机程序执行以下步骤：
S1，获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；
S2，获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；
S3，根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重。
可选地，在本实施例中，上述存储介质可以包括但不限于：U盘、只读存储器(Read-Only Memory，简称为ROM)、随机存取存储器(Random Access Memory，简称为RAM)、移动硬盘、磁碟或者光盘等各种可以存储程序代码的介质。
可选地，本实施例中的具体示例可以参考上述实施例及可选实施方式中所描述的示例，本实施例在此不再赘述。
显然，本领域的技术人员应该明白，上述的本公开的各模块或各步骤可以用通用的计算装置来实现，它们可以集中在单个的计算装置上，或者分布在多个计算装置所组成的网络上，可选地，它们可以用计算装置可执行的程序代码来实现，从而，可以将它们存储在存储装置中由计算装置来执行，并且在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤，或者将它们分别制作成各个集成电路模块，或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。这样，本公开不限制于任何特定的硬件和软件结合。
以上所述仅为本公开的优选实施例而已，并不用于限制本公开，对于本领域的技术人员来说，本公开可以有各种更改和变化。凡在本公开的原则之内，所作的任何修改、等同替换、改进等，均应包含在本公开的保护范围之内。

Claims (16)

  1. 一种权重的调整方法，包括：
    获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；
    获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；
    根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重。
  2. 根据权利要求1所述的权重的调整方法，其中，根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重，包括：
    确定所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作；
    根据所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作调整所述多个第二连续动作分别对应的多个第一权重。
  3. 根据权利要求2所述的权重的调整方法，其中，根据所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作调整所述多个第二连续动作分别对应的多个第一权重，包括：
    在所述多个第二连续动作中存在与所述第一连续动作全部一致的第三连续动作的情况下，将所述第三连续动作的第一权重增加至第二权重；
    在所述多个第二连续动作中存在与所述第一连续动作部分一致的第三连续动作的情况下，将所述第三连续动作的第一权重增加至第二权 重，以及将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重；
    在所述多个第二连续动作中存在与所述第一连续动作全部和部分一致的多个第三连续动作的情况下，分别将所述多个第三连续动作分别对应的第一权重增加至第二权重。
  4. 根据权利要求2所述的权重的调整方法，其中，根据所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作调整所述多个第二连续动作分别对应的多个第一权重，包括：
    在所述多个第二连续动作中不存在与所述第一连续动作全部一致的第三连续动作的情况下，将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重；
    在所述多个第二连续动作中不存在与所述第一连续动作部分一致的第三连续动作的情况下，不调整所述多个第二连续动作分别对应的多个第一权重；
    在所述多个第二连续动作中不存在与所述第一连续动作全部和部分一致的第三连续动作的情况下，将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重，以及不调整所述多个第二连续动作分别对应的多个第一权重。
  5. 根据权利要求1所述的权重的调整方法，其中，根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重之后，所述方法还包括：
    确定所述行为数据集是否到达衰减周期；
    在所述行为数据集到达衰减周期的情况下，获取当前时刻、上一次衰减时刻；
    根据所述当前时刻、所述上一次衰减时刻确定所述行为数据集的衰 减因子；
    根据所述衰减因子调整所述行为数据集中的多个第二连续动作的第一权重。
  6. 根据权利要求5所述的权重的调整方法，其中，根据所述衰减因子调整所述行为数据集中的多个第二连续动作的第一权重，包括：
    根据以下公式确定所述行为数据集的衰减因子：
    Figure PCTCN2022100559-appb-100001
    其中，t为当前时刻、t0为上一次衰减时刻，α为可变参数；
    将所述衰减因子乘调整后的第一权重，以得到所述行为数据集中的多个第二连续动作的第三权重。
  7. 根据权利要求1所述的权重的调整方法，其中，根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重之后，所述方法还包括：
    获取所述目标对象的第一动作，其中，所述第一动作为用于指示所述目标对象的单独动作；
    在所述目标数据集中确定包含所述第一动作的第四连续动作和所述第四连续动作对应的第四权重；
    根据所述第四连续动作和所述第四权重预测所述目标对象的待执行动作。
  8. 一种权重的调整装置，包括：
    第一获取模块，设置为获取目标对象预设时间段内的第一连续动作，其中，所述预设时间段为当前时刻之前的时间段；
    第二获取模块，设置为获取所述目标对象的行为数据集，其中，所述行为数据集包括：多个第二连续动作，以及多个第二连续动作分别对应的多个第一权重；所述第二连续动作与所述第一权重一一对应；
    调整模块，设置为根据所述第一连续动作调整所述多个第二连续动作分别对应的多个第一权重。
  9. 根据权利要求8所述的权重的调整装置，其中，
    所述调整模块，还设置为确定所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作；根据所述多个第二连续动作中是否存在与所述第一连续动作全部和/或部分一致的第三连续动作调整所述多个第二连续动作分别对应的多个第一权重。
  10. 根据权利要求9所述的权重的调整装置，其中，
    所述调整模块76，还设置为在所述多个第二连续动作中存在与所述第一连续动作全部一致的第三连续动作的情况下，将所述第三连续动作的第一权重增加至第二权重；在所述多个第二连续动作中存在与所述第一连续动作部分一致的第三连续动作的情况下，将所述第三连续动作的第一权重增加至第二权重，以及将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重；在所述多个第二连续动作中存在与所述第一连续动作全部和部分一致的多个第三连续动作的情况下，分别将所述多个第三连续动作分别对应的第一权重增加至第二权重。
  11. 根据权利要求9所述的权重的调整装置，其中，
    所述调整模块，还设置为在所述多个第二连续动作中不存在与所述第一连续动作全部一致的第三连续动作的情况下，将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重；在所述多个第二连续动作中不存在与所述第一连续动作部分一致的第三连续动作的情况下，不调整所述多个第二连续动作分别对应的多个第一权重；在所述多个第二连续动作中不存在与所述第一连续动作全部和部分一致的第三连续动作的情况下，将所述第一连续动作添加至所述行为数据集，并为所述第一连续动作赋予默认权重，以及不调整所述多个第二连 续动作分别对应的多个第一权重。
  12. 根据权利要求8所述的权重的调整装置，其中，
    所述调整模块，还设置为确定所述行为数据集是否到达衰减周期；在所述行为数据集到达衰减周期的情况下，获取当前时刻、上一次衰减时刻；根据所述当前时刻、所述上一次衰减时刻确定所述行为数据集的衰减因子；根据所述衰减因子调整所述行为数据集中的多个第二连续动作的第一权重。
  13. 根据权利要求12所述的权重的调整装置，其中，
    所述调整模块76，还设置为根据以下公式确定所述行为数据集的衰减因子：
    Figure PCTCN2022100559-appb-100002
    其中，t为当前时刻、t0为上一次衰减时刻，α为可变参数；将所述衰减因子乘调整后的第一权重，以得到所述行为数据集中的多个第二连续动作的第三权重。
  14. 根据权利要求8所述的权重的调整装置，其中，
    所述第一获取模块，还设置为获取所述目标对象的第一动作，其中，所述第一动作为用于指示所述目标对象的单独动作；在所述目标数据集中确定包含所述第一动作的第四连续动作和所述第四连续动作对应的第四权重；根据所述第四连续动作和所述第四权重预测所述目标对象的待执行动作。
  15. 一种计算机可读的存储介质，所述计算机可读的存储介质包括存储的程序，其中，所述程序运行时执行上述权利要求1至7任一项中所述的方法。
  16. 一种电子装置，包括存储器和处理器，所述存储器中存储有计算机程序，所述处理器被设置为通过所述计算机程序执行所述权利要求1至7任一项中所述的方法。
PCT/CN2022/100559 2022-01-28 2022-06-22 权重的调整方法和装置、存储介质及电子装置 WO2023142351A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210109432.0A CN114399057A (zh) 2022-01-28 2022-01-28 权重的调整方法和装置、存储介质及电子装置
CN202210109432.0 2022-01-28

Publications (1)

Publication Number Publication Date
WO2023142351A1 true WO2023142351A1 (zh) 2023-08-03

Family

ID=81232278

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/100559 WO2023142351A1 (zh) 2022-01-28 2022-06-22 权重的调整方法和装置、存储介质及电子装置

Country Status (2)

Country Link
CN (1) CN114399057A (zh)
WO (1) WO2023142351A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114399057A (zh) * 2022-01-28 2022-04-26 青岛海尔科技有限公司 权重的调整方法和装置、存储介质及电子装置

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111528135A (zh) * 2020-04-15 2020-08-14 上海明略人工智能(集团)有限公司 目标对象的确定方法及装置、存储介质、电子装置
CN112030465A (zh) * 2020-08-14 2020-12-04 海尔优家智能科技（北京）有限公司 第一对象的清洗方法及装置、存储介质、电子装置
CN112613642A (zh) * 2020-12-07 2021-04-06 国网北京市电力公司 应急物资需求预测方法和装置、存储介质及电子设备
US20210133569A1 (en) * 2019-11-04 2021-05-06 Tsinghua University Methods, computing devices, and storage media for predicting traffic matrix
CN113205370A (zh) * 2021-05-27 2021-08-03 北京深演智能科技股份有限公司 数据处理方法、数据处理装置及电子设备
CN114399057A (zh) * 2022-01-28 2022-04-26 青岛海尔科技有限公司 权重的调整方法和装置、存储介质及电子装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20210133569A1 (en) * 2019-11-04 2021-05-06 Tsinghua University Methods, computing devices, and storage media for predicting traffic matrix
CN111528135A (zh) * 2020-04-15 2020-08-14 上海明略人工智能(集团)有限公司 目标对象的确定方法及装置、存储介质、电子装置
CN112030465A (zh) * 2020-08-14 2020-12-04 海尔优家智能科技（北京）有限公司 第一对象的清洗方法及装置、存储介质、电子装置
CN112613642A (zh) * 2020-12-07 2021-04-06 国网北京市电力公司 应急物资需求预测方法和装置、存储介质及电子设备
CN113205370A (zh) * 2021-05-27 2021-08-03 北京深演智能科技股份有限公司 数据处理方法、数据处理装置及电子设备
CN114399057A (zh) * 2022-01-28 2022-04-26 青岛海尔科技有限公司 权重的调整方法和装置、存储介质及电子装置

Also Published As

Publication number Publication date
CN114399057A (zh) 2022-04-26

Similar Documents

Publication Publication Date Title
KR102339239B1 (ko) 클라우드-디바이스 협업적 실시간 사용자 사용 및 성능 비정상 검출을 위한 시스템 및 방법
Apostolopoulos et al. Cognitive data offloading in mobile edge computing for internet of things
US10873864B2 (en) Incorporation of expert knowledge into machine learning based wireless optimization framework
CN113315716A (zh) 拥塞控制模型的训练方法和设备及拥塞控制方法和设备
EP3931698A1 (en) Dynamic network configuration
CN105407158A (zh) 一种建立模型和推送消息的方法及装置
WO2023142351A1 (zh) 权重的调整方法和装置、存储介质及电子装置
US20230004776A1 (en) Moderator for identifying deficient nodes in federated learning
CN111935025A (zh) 一种tcp传输性能的控制方法、装置、设备和介质
CN114726799A (zh) 拥塞控制智能体的训练方法、拥塞控制方法及装置
US20230216737A1 (en) Network performance assessment
CN114090108A (zh) 算力任务执行方法、装置、电子设备及存储介质
CN112494935A (zh) 一种云游戏平台池化方法、电子设备及存储介质
CN110874635A (zh) 一种深度神经网络模型压缩方法及装置
CN111046156B (zh) 奖励数据的确定方法、装置和服务器
WO2023159821A1 (zh) 操作行为的确定方法和装置、存储介质及电子装置
CN113543160A (zh) 5g切片资源配置方法、装置、计算设备及计算机存储介质
CN112669091B (zh) 数据处理方法、装置及存储介质
CN113055423B (zh) 策略的推送方法、策略的执行方法、装置、设备及介质
US10931537B2 (en) Method and device for updating a predictive model of a variable relating to a mobile terminal
CN117172627B (zh) 一种业务执行方法、装置、设备及存储介质
Zhang et al. Adaptive Frame Rate Optimization Based on Particle Swarm and Neural Network for Industrial Video Stream
CN113055212A (zh) 策略的推送方法、策略的执行方法、装置、设备及介质
CN112702276B (zh) 一种传输控制方法、装置、电子设备及计算机存储介质
JP7478300B1 (ja) 通信制御装置および通信制御方法

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: 22923160

Country of ref document: EP

Kind code of ref document: A1