CN114357499A - Data management method and system based on joint learning - Google Patents
Data management method and system based on joint learning Download PDFInfo
- Publication number
- CN114357499A CN114357499A CN202011095825.8A CN202011095825A CN114357499A CN 114357499 A CN114357499 A CN 114357499A CN 202011095825 A CN202011095825 A CN 202011095825A CN 114357499 A CN114357499 A CN 114357499A
- Authority
- CN
- China
- Prior art keywords
- data
- target
- model
- joint learning
- joint
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention is suitable for the technical field of cloud computing, and provides a data management method and a system based on joint learning, wherein the method comprises the following steps: acquiring target data, wherein the target data are non-shared data; carrying out local target model training on the target data to obtain at least one class of target models; establishing a joint learning global model according to the target model; and uploading the joint learning global model to demand equipment. According to the method, each user carries out model training locally in a joint learning mode, model parameters are uploaded to the cloud for joint training, and then a series of joint learning global models are generated. On the premise of not sharing local business data, a health guidance model with better effect can be obtained. With the increase of the joint learning global models, the models are gradually deposited in the cloud end and are provided for users of the life health ecosphere in a uniform standard interface mode, and health data sharing is achieved.
Description
Technical Field
The invention belongs to the technical field of cloud computing, and particularly relates to a data management method and system based on joint learning.
Background
Cloud health, also called health cloud, is an idea recently proposed, and refers to providing online, real-time and up-to-date services and derivative product development such as health management, disease treatment, disease diagnosis, human body function data acquisition and the like for medical patients and health demanders through the technical means of cloud computing, cloud storage, cloud service, internet of things, mobile internet and the like and through the association, interaction, communication and cooperation of relevant departments such as medical institutions, experts, medical research institutions, medical manufacturers and the like.
Cloud health is a system engineering and complex huge system across different industries such as electronics, communication, medical treatment, biology, software and the like, and needs government guidance and entrance and support of related industries. Medical internet of things equipment, digital hospitals, remote diagnosis, family intelligent doctors, intelligent medical treatment, electronic health files and the like applied in China all become important components of the medical internet of things equipment, the digital hospitals, the remote diagnosis, the family intelligent doctors, the intelligent medical treatment and the electronic health files.
The cloud health becomes the health management from the cradle to the grave, namely, the whole-process health management system of the whole person can record the pathogenic causes of the human body by accurate medical terms, and doctors can find problems conveniently. From the birth examination record of the fetal period, to the daily physical examination report, the real-time human body physical sign data and the inquiry result of each doctor, each information related to health is recorded in the form of data, and community responsible doctors can help residents to carry out health management according to the records, remind the residents of paying attention to health matters, and meanwhile, the gathering of massive resident health information can also help disease control departments to carry out local epidemiological statistics, discover high-incidence diseases in various regions and carry out high-incidence prevention and treatment work. The realization and popularization of cloud health in the future can greatly improve the diagnosis efficiency and the treatment level and inject more beneficial factors for human health.
However, the user health data includes health profile data, user profile data, disease diagnosis data, health intervention data, and the like of the user. Each kind of data is sensitive information extremely private to users, any leakage or random sharing is carried out, even the data is illegal, and therefore the application appeal of the joint learning is achieved. However, no scheme is available for sharing data and avoiding leakage of user privacy sensitive health data.
Disclosure of Invention
In view of this, embodiments of the present invention provide a data management method and system based on joint learning, so as to solve the problem that user health data is extremely private sensitive information of a user, and any leakage or random sharing, even all illegal, but there is no scheme that can implement data sharing and avoid leakage of user private sensitive health data.
A first aspect of an embodiment of the present invention provides a data management method based on joint learning, including:
s1, acquiring target data, wherein the target data are non-shared data;
s2, carrying out local target model training on the target data to obtain at least one type of target model;
s3, establishing a joint learning global model according to the target model;
and S4, uploading the joint learning global model to a demand device.
Further, the acquiring the target data includes:
responding to target equipment, and acquiring attribute information of the target equipment;
establishing a simulation database based on joint learning according to the attribute information;
acquiring the target data by using the simulation database;
the target data is health data.
Further, the performing local target model training on the target data to obtain at least one class of target models includes:
acquiring corresponding local data according to the target data;
classifying the local data;
and performing combined learning model training according to the classification to obtain at least one class of target models.
Further, establishing a joint learning global model according to the target model, specifically: and averaging or weighted averaging parameters in the local model through a joint learning engine framework at the cloud end according to the target model, and establishing a joint learning global model.
Further, after uploading the joint learning global model to a demand device, the method further includes: and (3) continuously adding and updating new user data in the joint learning global model, carrying out local target model training on the updated user data, simultaneously carrying out average or weighted average on the parameters of the trained new local model, continuously and iteratively establishing a new joint learning global model, and finally, depositing the joint learning global model on the cloud end and uploading the deposit to a demand device.
A second aspect of an embodiment of the present invention provides a data management system based on joint learning, including:
the system comprises a target data acquisition module, a health data acquisition end and a data processing module, wherein the target data acquisition module is used for acquiring target data through the health data acquisition end, and the target data is non-shared data;
the local target model training module is used for carrying out local target model training on the target data to obtain at least one class of target models;
the global model establishing module is used for establishing a joint learning global model according to the target model;
and the global model uploading module is used for uploading the joint learning global model to the demand equipment.
Further, the acquiring of the target data by the target data acquiring module includes:
responding to target equipment, and acquiring attribute information of the target equipment;
establishing a simulation database based on joint learning according to the attribute information;
acquiring the target data by using the simulation database;
the target data is health data.
Further, the global model establishing module establishes a joint learning global model according to the target model, specifically: and averaging or weighted averaging parameters in the local model through a joint learning engine framework at the cloud end according to the target model, and establishing a joint learning global model.
A third aspect of the embodiments of the present invention provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of the joint learning-based data management method when executing the computer program.
A fourth aspect of embodiments of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the joint learning-based data management method as described above.
Compared with the prior art, the embodiment of the invention has the following beneficial effects:
according to the method, each user carries out model training locally in a joint learning mode, model parameters are uploaded to the cloud for joint training, and then a series of joint learning global models are generated. On the premise of not sharing local business data, a health guidance model with better effect can be obtained. With the increase of the joint learning global models, the models are gradually deposited in the cloud end and are provided for users of the life health ecosphere in a uniform standard interface mode, and health data sharing is achieved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed for the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
FIG. 1 is a detailed architecture diagram of a joint learning-based data management system according to an embodiment of the present invention;
FIG. 2 is a schematic block diagram of a system provided by an embodiment of the present invention;
FIG. 3 is a flow chart of a method provided by an embodiment of the present invention;
FIG. 4 is an exemplary diagram of a joint learning application provided by embodiments of the present invention;
fig. 5 is a schematic diagram of a terminal device according to an embodiment of the present invention.
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.
In order to explain the technical means of the present invention, the following description will be given by way of specific examples.
Example 1
As shown in fig. 1, 3 and 4, the present embodiment provides a data management method based on joint learning, including:
and S1, acquiring target data, wherein the target data is non-shared data. The method for acquiring the target data comprises the following steps: firstly, responding to target equipment (such as a sphygmomanometer, a blood glucose meter and a body fat monitor) to obtain attribute information of the target equipment; then, establishing a simulation database based on joint learning according to the attribute information; finally, the simulation database is used for obtaining the target data, the target data is life health class data and at least comprises the following steps: the information data of the user body at least comprises health file data, client portrait data, diagnosis data, intervention data and life style data. As shown in fig. 4, wherein:
health profile data, comprising: 1. personal information such as name, gender, address, age, etc.; 2. historical health data, such as disease history, allergy history physical examination data; 3. three-therapy detection, namely information data (including mood index, pressure data, depression data and the like) of the body of a user, energy data (including meridian data, movement data, metabolism data and the like), biomedical detection data (including body temperature, blood pressure, blood fat, blood sugar and the like)
Customer representation data, such as consumption preference data, implemented through keyword searches; consumption requirement data such as browsing preferences, interest preferences.
Diagnostic data such as triple therapy reports (biomedical diagnostic reports, energy medical diagnostic reports, information medical diagnostic reports).
Intervention data, such as triple therapy intervention data, behavioral change data, conditioning data, process visualization data. Hospital intervention data, such as physiological monitoring data, medication, treatment data, care data.
Lifestyle data, such as living environment data, diet data, sleep data, exercise data.
As shown in fig. 1, looking from bottom to top in the figure, the bottom is various human health data monitoring terminals, for example, the user terminal monitoring device can be an intelligent health mirror, an intelligent toilet, an intelligent environment box, a sleep monitor, a risk assessment scale, an intelligent seat cushion, a health bracelet, a sphygmomanometer, a blood glucose meter, a body fat monitor, an electrocardiogram monitor and an oximeter. There are also some community health monitoring devices, such as physical examination instruments, electroencephalogram, quantum micromagnetometers. The life health equipment uniformly accesses data to a data platform according to a certain Internet of things standard (CIM standard), and the data is transmitted to an upper data platform through a 4G gateway, a universal energy box and an XCM soft gateway.
S2, performing local target model training on the target data to obtain at least one category of target model, specifically, performing local target model training on the target data in a joint learning manner, obtaining corresponding local data (data acquired locally in real time, such as blood glucose value and blood pressure value) according to the target data, classifying the local data, and performing joint learning model training according to the classification to obtain at least one category of target model. As shown in fig. 1, when the data center performs local model training on target data, the following data services are provided for health data, including: health data standards and data management, including at least biomedical, energy medicine, information medicine, knowledge image base; big data calculation at least comprising real-time calculation, off-line calculation, data warehouse and data visualization; and managing the data service, which at least comprises service registration, service monitoring, service discovery and service governance.
And S3, establishing a joint learning global model according to the target model, specifically, establishing the joint learning global model through a joint learning engine framework at the cloud, gradually depositing the joint learning global model on the cloud as new user data is added and is uploaded to the demand equipment. As shown in fig. 1, the present embodiment provides a joint learning global model applied in the life health industry, which includes a health assessment model, a health monitoring model, a health recommendation model, a disease early warning model, a knowledge graph model, and a personalized intervention model. The joint learning engine comprises joint learning framework design, distributed exception handling, edge cloud cooperative communication protocols, joint strategies, security strategies, aggregation strategies and engine control, and management of communication logic and communication protocols is achieved through the joint learning engine.
And S4, uploading the joint learning global model to the demand equipment. Various basic statistical algorithms, machine learning algorithms and deep learning algorithms are adopted when the local model training and the joint learning global model building are carried out on the target, the basic statistical algorithms comprise image recognition, video analysis and image understanding, the machine learning algorithms comprise classification, regression and clustering, and the deep learning algorithms comprise CNN and RNN deep learning algorithms; and packaging a series of algorithm service APIs based on the joint learning global model to enable the user to interact with the joint learning engine, and realizing business scene application by using the joint learning engine.
In addition, the algorithm component module is adopted when the local model training is carried out on the target and the joint learning global model is established, and comprises a basic model training component and an applicable model component, wherein the basic model training component comprises a data loading module, a general model training module and a model evaluation component, and the applicable model component comprises an intelligent diagnosis model component, an intelligent detection type component and an intelligent intervention model component.
The server can not collect data but can collect parameters of the model in the joint learning process, and the server coordinates the edge devices to participate in training, wherein each edge device has training data. Each edge device trains a local model by using own data, encrypts and uploads own parameters to a server, and the server averages or weights the collected parameters to form a joint learning global model applied in the life health industry. Broadcast to each edge device. For example, the current model is downloaded by the edge device and then trained with the data on the handset, after which all changes are summarized as a small update. Finally, only this update is passed to the cloud (using encrypted communication) and immediately averaged with the other user's updates, which then improves the shared model. All training data are kept on the edge device, the training data and the data model are not local, and the cloud end cannot store independent updates.
As shown in fig. 4, the working principle of the present embodiment is:
from bottom to top, the user health data, including the user's health profile data, user profile data, disease diagnosis data, health intervention data, etc., is below. Each type of data is sensitive information that is extremely private to the user, and any leakage or random sharing is even illegal. This has the application appeal of joint learning
In order to obtain a health guidance model with a better effect on the premise of not sharing local business data, each user carries out model training locally in a joint learning mode, model parameters are uploaded to a cloud end for joint training, and then a series of global models for joint learning are generated.
With the increase of the joint learning global models, the models are gradually deposited in the cloud end and are provided for partners of the life health ecosphere in a uniform standard interface mode.
Example 2
As shown in fig. 1 and 2, the present embodiment provides a data management system based on joint learning, including:
and a target data acquisition module 101 (i.e. a middle station of the internet of things) arranged on the Iaas layer, configured to acquire target data, where the target data is non-shared data. The target data is life health class data, and at least comprises the following data: information data of the body of the user at least comprising mood index, stress data and depression data; energy data at least comprising meridian data, movement data and metabolism data; the user community data at least comprises community physical examination data and community living center data.
As shown in fig. 1, looking from bottom to top in the figure, the bottom is various human health data monitoring terminals, for example, the user terminal monitoring device can be an intelligent health mirror, an intelligent toilet, an intelligent environment box, a sleep monitor, a risk assessment scale, an intelligent seat cushion, a health bracelet, a sphygmomanometer, a blood glucose meter, a body fat monitor, an electrocardiogram monitor and an oximeter. There are also some community health monitoring devices, such as physical examination instruments, electroencephalogram, quantum micromagnetometers. The life health equipment uniformly accesses data to a data platform according to a certain Internet of things standard (CIM standard), and the data is transmitted to an upper data platform through a 4G gateway, a universal energy box and an XCM soft gateway.
The system comprises a local target model training module 102 (namely, a data middlebox) and a global model establishing module 103 (namely, an artificial intelligence middlebox) which are arranged on a Paas layer, wherein the local target model training module 102 is used for performing local target model training on target data to obtain at least one category of target model, and specifically is used for performing local target model training on the target data in a joint learning mode. As shown in fig. 1, when the data center performs local model training on target data, the following data services are provided for health data, including: health data standards and data management, including at least biomedical, energy medicine, information medicine, knowledge image base; big data calculation at least comprising real-time calculation, off-line calculation, data warehouse and data visualization; data service management, at least comprising service registration, service monitoring, service discovery and service management;
the global model building module 103 (i.e., an artificial intelligence middle desk) is configured to build a joint learning global model according to the target model, specifically, build the joint learning global model through a joint learning engine framework at the cloud, gradually deposit the joint learning global model on the cloud as new user data is added more and more, and upload the deposited joint learning global model to a demand device. As shown in fig. 1, the present embodiment provides a joint learning global model applied in the life health industry, which includes a health assessment model, a health monitoring model, a health recommendation model, a disease early warning model, a knowledge graph model, and a personalized intervention model. The joint learning engine comprises joint learning framework design, distributed exception handling, edge cloud cooperative communication protocols, joint strategies, security strategies, aggregation strategies and engine control, and management of communication logic and communication protocols is achieved through the joint learning engine.
And the global model uploading module 104 (namely, the life health ecological client) arranged at the Saas layer is used for uploading the joint learning global model to the demand equipment, interacting with the joint learning engine, and realizing business scene application by using the joint learning engine.
Wherein, IaaS, PaaS and SaaS are three service modes of cloud computing, and the specific analysis is as follows:
1, SaaS: Software-as-a-Service (softas-as-a-Service) provides services to customers that are applications run by operators on cloud computing infrastructure that users can access through client interfaces, such as browsers, on a variety of devices. The consumer does not need to manage or control any cloud computing infrastructure, including networks, servers, operating systems, storage, and the like;
PaaS: Platform-as-a-Service (Platform as a Service) provides a Service to consumers by deploying applications developed or purchased by customers using a provided development language and tools (e.g., Java, python,. Net, etc.) to a provider's cloud computing infrastructure. The customer does not need to manage or control the underlying cloud infrastructure, including networks, servers, operating systems, storage, etc., but can control deployed applications and possibly also the configuration of the hosting environment in which the applications are run;
IaaS: the services provided by Infrastructure-as-a-Service to consumers are a utilization of all the computing Infrastructure, including processing CPU, memory, storage, network and other basic computing resources, that users can deploy and run arbitrary software, including operating systems and applications. Consumers do not manage or control any cloud computing infrastructure, but can control operating system selection, storage space, deployed applications, and possibly limited network components (e.g., routers, firewalls, load balancers, etc.).
Meanwhile, various basic statistical algorithms, machine learning algorithms and deep learning algorithms are adopted when the local model training and the joint learning global model building are carried out on the target, the basic statistical algorithms comprise image recognition, video analysis and image understanding, the machine learning algorithms comprise classification, regression and clustering, and the deep learning algorithms comprise CNN and RNN deep learning algorithms; and packaging a series of algorithm service APIs based on the joint learning global model to enable the user to interact with the joint learning engine, and realizing business scene application by using the joint learning engine.
In addition, the algorithm component module is adopted when the local model training is carried out on the target and the joint learning global model is established, and comprises a basic model training component and an applicable model component, wherein the basic model training component comprises a data loading module, a general model training module and a model evaluation component, and the applicable model component comprises an intelligent diagnosis model component, an intelligent detection type component and an intelligent intervention model component.
The server can not collect data but can collect parameters of the model in the joint learning process, and the server coordinates the edge devices to participate in training, wherein each edge device has training data. Each edge device trains a local model by using own data, encrypts and uploads own parameters to a server, and the server averages or weights the collected parameters to form a joint learning global model applied in the life health industry. Broadcast to each edge device. For example, the current model is downloaded by the edge device and then trained with the data on the handset, after which all changes are summarized as a small update. Finally, only this update is passed to the cloud (using encrypted communication) and immediately averaged with the other user's updates, which then improves the shared model. All training data are kept on the edge device, the training data and the data model are not local, and the cloud end cannot store independent updates.
As shown in fig. 4, the working principle of the present embodiment is:
from bottom to top, the user health data, including the user's health profile data, user profile data, disease diagnosis data, health intervention data, etc., is below. Each type of data is sensitive information that is extremely private to the user, and any leakage or random sharing is even illegal. This has the application appeal of joint learning
In order to obtain a health guidance model with a better effect on the premise of not sharing local business data, each user carries out model training locally in a joint learning mode, model parameters are uploaded to a cloud end for joint training, and then a series of global models for joint learning are generated.
With the increase of the joint learning global models, the models are gradually deposited in the cloud end and are provided for partners of the life health ecosphere in a uniform standard interface mode.
It should be understood that, the sequence numbers of the steps in the foregoing embodiments do not imply an execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.
Fig. 5 is a schematic diagram of a terminal device according to an embodiment of the present invention. As shown in fig. 5, the terminal device 6 of this embodiment includes: a processor 60, a memory 61, and a computer program 62, such as a joint learning training program, stored in the memory 61 and executable on the processor 60. The processor 60, when executing the computer program 62, implements the steps in the various joint learning based data management method embodiments described above, such as the steps 101 to 104 shown in fig. 1. Alternatively, the processor 60, when executing the computer program 62, implements the functions of the various modules/units in the above-described apparatus embodiments, such as the functions of the modules 51 to 54 shown in fig. 5.
Illustratively, the computer program 62 may be divided into one or more modules/units, which are stored in the memory 61 and executed by the processor 60 to implement the present invention. One or more of the modules/units may be a series of computer program instruction segments capable of performing specific functions, which are used to describe the execution of the computer program 62 in the terminal device 6. For example, the computer program 62 may be divided into a synchronization module, a summarization module, an acquisition module, and a return module (a module in a virtual device), each of which functions specifically as follows:
the terminal device 6 may be a desktop computer, a notebook, a palm computer, a cloud server, or other computing devices. The terminal device may include, but is not limited to, a processor 60, a memory 61. Those skilled in the art will appreciate that fig. 5 is merely an example of a terminal device 6 and does not constitute a limitation of terminal device 6 and may include more or fewer components than shown, or some components may be combined, or different components, e.g., the terminal device may also include input output devices, network access devices, buses, etc.
The Processor 60 may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The storage 61 may be an internal storage unit of the terminal device 6, such as a hard disk or a memory of the terminal device 6. The memory 61 may also be an external storage device of the terminal device 6, such as a plug-in hard disk provided on the terminal device 6, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like. Further, the memory 61 may also include both an internal storage unit of the terminal device 6 and an external storage device. The memory 61 is used for storing computer programs and other programs and data required by the terminal device. The memory 61 may also be used to temporarily store data that has been output or is to be output.
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-mentioned division of the functional units and modules is illustrated, and in practical applications, the above-mentioned function distribution may be performed by different functional units and modules according to needs, that is, the internal structure of the apparatus is divided into different functional units or modules, so as to perform all or part of the functions described above. Each functional unit and module in the embodiments may be integrated in one processing unit, or each unit may exist alone physically, or two or more units are integrated in one unit, and the integrated unit may be implemented in a form of hardware, or in a form of software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working processes of the units and modules in the system may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and reference may be made to the related descriptions of other embodiments for parts that are not described or illustrated in a certain embodiment.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus/terminal device and method may be implemented in other ways. For example, the above-described embodiments of the apparatus/terminal device are merely illustrative, and for example, the division of the modules or units is only one logical division, and there may be other divisions when actually implemented, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated modules/units, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium. Based on such understanding, all or part of the flow of the method according to the embodiments of the present invention may also be implemented by a computer program, which may be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method embodiments may be implemented. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, and the like. It should be noted that the computer readable medium may contain content that is subject to appropriate increase or decrease as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer readable media does not include electrical carrier signals and telecommunications signals as is required by legislation and patent practice.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present invention, and are intended to be included within the scope of the present invention.
Claims (10)
1. A data management method based on joint learning is characterized by comprising the following steps:
s1, acquiring target data, wherein the target data are non-shared data;
s2, carrying out local target model training on the target data to obtain at least one type of target model;
s3, establishing a joint learning global model according to the target model;
and S4, uploading the joint learning global model to a demand device.
2. The joint learning-based data management method according to claim 1, wherein the acquiring target data includes:
responding to target equipment, and acquiring attribute information of the target equipment;
establishing a simulation database based on joint learning according to the attribute information;
acquiring the target data by using the simulation database;
the target data is health data.
3. The method for data management based on joint learning according to claim 1, wherein the local target model training of the target data to obtain at least one category of target model comprises:
acquiring corresponding local data according to the target data;
classifying the local data;
and performing combined learning model training according to the classification to obtain at least one class of target models.
4. The joint learning-based data management method according to claim 1, wherein the joint learning global model is established according to the target model, and specifically comprises: and averaging or weighted averaging parameters in the local model through a joint learning engine framework at the cloud end according to the target model, and establishing a joint learning global model.
5. The joint learning-based data management method according to claim 1, wherein after uploading the joint learning global model to a demand device, the method further comprises: and (3) continuously adding and updating new user data in the joint learning global model, carrying out local target model training on the updated user data, simultaneously carrying out average or weighted average on the parameters of the trained new local model, continuously and iteratively establishing a new joint learning global model, and finally, depositing the joint learning global model on the cloud end and uploading the deposit to a demand device.
6. A joint learning-based data management system, comprising:
the system comprises a target data acquisition module, a health data acquisition end and a data processing module, wherein the target data acquisition module is used for acquiring target data through the health data acquisition end, and the target data is non-shared data;
the local target model training module is used for carrying out local target model training on the target data to obtain at least one class of target models;
the global model establishing module is used for establishing a joint learning global model according to the target model;
and the global model uploading module is used for uploading the joint learning global model to the demand equipment.
7. The joint learning-based data management system according to claim 6, wherein the target data acquisition module acquires target data includes:
responding to target equipment, and acquiring attribute information of the target equipment;
establishing a simulation database based on joint learning according to the attribute information;
acquiring the target data by using the simulation database;
the target data is health data.
8. The joint learning-based data management system according to claim 6, wherein the global model building module builds a joint learning global model according to the target model, specifically: and averaging or weighted averaging parameters in the local model through a joint learning engine framework at the cloud end according to the target model, and establishing a joint learning global model.
9. A terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the joint learning-based data management method according to any one of claims 1 to 5 when executing the computer program.
10. A computer-readable storage medium, in which a computer program is stored, which, when being executed by a processor, carries out the steps of the joint learning-based data management method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011095825.8A CN114357499A (en) | 2020-10-14 | 2020-10-14 | Data management method and system based on joint learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011095825.8A CN114357499A (en) | 2020-10-14 | 2020-10-14 | Data management method and system based on joint learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114357499A true CN114357499A (en) | 2022-04-15 |
Family
ID=81089790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011095825.8A Pending CN114357499A (en) | 2020-10-14 | 2020-10-14 | Data management method and system based on joint learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114357499A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116805175A (en) * | 2023-06-02 | 2023-09-26 | 中哲国际工程设计有限公司 | Medical care building operation and maintenance management system based on CIM technology |
-
2020
- 2020-10-14 CN CN202011095825.8A patent/CN114357499A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116805175A (en) * | 2023-06-02 | 2023-09-26 | 中哲国际工程设计有限公司 | Medical care building operation and maintenance management system based on CIM technology |
| CN116805175B (en) * | 2023-06-02 | 2023-12-26 | 中哲国际工程设计有限公司 | Medical care building operation and maintenance management system based on CIM technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Philip et al. | Internet of Things for in-home health monitoring systems: Current advances, challenges and future directions | |
| Al-Khafajiy et al. | Remote health monitoring of elderly through wearable sensors | |
| Farahani et al. | Healthcare iot | |
| Volkov et al. | Digital twins, internet of things and mobile medicine: a review of current platforms to support smart healthcare | |
| Ma et al. | Intelligent healthcare systems assisted by data analytics and mobile computing | |
| Estrela et al. | Health 4.0: applications, management, technologies and review | |
| Miah et al. | On-cloud healthcare clinic: an e-health consultancy approach for remote communities in a developing country | |
| CN112133445A (en) | Cardiovascular disease management service method and system | |
| Jain et al. | Adoption of fog computing in healthcare 4.0 | |
| JP2017517088A (en) | Method, apparatus and system for improving health condition | |
| CN108428477A (en) | Construction method and cloud medical system based on the twinborn cloud surgery simulation platform of number | |
| CN103690240A (en) | Medical system | |
| Gupta et al. | Healthcare 4.0: recent advancements and futuristic research directions | |
| Mishra et al. | A novel approach towards using big data and IoT for improving the efficiency of m-health systems | |
| Rahman et al. | An IoT and blockchain-based multi-sensory in-home quality of life framework for cancer patients | |
| CN112447299A (en) | Medical care resource prediction model training method, device, equipment and storage medium | |
| Bagga et al. | Applications of mobile agents in healthcare domain: A literature survey | |
| Gowsalya et al. | Predicting the risk of readmission of diabetic patients using MapReduce | |
| CN111986812A (en) | Internet-based gynecological disease medical diagnosis and treatment control system | |
| Dash et al. | A comprehensive study of mobile computing in telemedicine | |
| CN114357499A (en) | Data management method and system based on joint learning | |
| Zhang et al. | Healthcare-based on cloud electrocardiogram system: a medical center experience in middle Taiwan | |
| Alahmadi et al. | A privacy-preserved IoMT-based mental stress detection framework with federated learning | |
| Israni et al. | Human‐Machine Interaction in Leveraging the Concept of Telemedicine | |
| Comito et al. | A power-aware approach for smart health monitoring and decision support |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |