CN115827224A - Multi-task one-way capsule network resource scheduling method based on federal learning - Google Patents
Multi-task one-way capsule network resource scheduling method based on federal learning Download PDFInfo
- Publication number
- CN115827224A CN115827224A CN202211438714.1A CN202211438714A CN115827224A CN 115827224 A CN115827224 A CN 115827224A CN 202211438714 A CN202211438714 A CN 202211438714A CN 115827224 A CN115827224 A CN 115827224A
- Authority
- CN
- China
- Prior art keywords
- task
- capsule
- communication
- cabin
- tasks
- 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
- 239000002775 capsule Substances 0.000 title claims abstract description 68
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000004891 communication Methods 0.000 claims abstract description 100
- 230000006854 communication Effects 0.000 claims abstract description 99
- 239000003999 initiator Substances 0.000 claims abstract description 7
- 230000004044 response Effects 0.000 claims abstract description 5
- 238000002360 preparation method Methods 0.000 claims description 4
- 239000003795 chemical substances by application Substances 0.000 claims description 3
- 238000004590 computer program Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 8
- 230000005540 biological transmission Effects 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000002457 bidirectional effect Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007175 bidirectional communication Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000001125 extrusion Methods 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000013024 troubleshooting Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/10—File systems; File servers
- G06F16/18—File system types
- G06F16/182—Distributed file systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Computer And Data Communications (AREA)
Abstract
The invention provides a multi-task one-way capsule network resource scheduling method based on federal learning, and relates to the technical field of federal learning modeling. The method comprises the following steps: the initiator initializes the task and the data source, and simultaneously, the virtual single communication coordination module synchronously informs the participant of starting a response. When multiple tasks are parallel, a data packet to be sent of any task is decomposed into a plurality of task communication packets. And placing the plurality of task communication packets into a waiting cabin of the capsule cabin according to the seat proportion table. When the waiting cabin loading is finished, the communication is coordinated according to the single communication coordination module. The capsule cabin starts multithreading to be sent to the participators, and the participators place the data in the capsule cabin into a task queue according to the tasks. And when any task acquires the waiting data, entering the next stage of the task to complete the task, and repeatedly executing the steps until all tasks are completed. The method enables a plurality of tasks to be completed smoothly in a controllable communication mode, and the condition of local communication blockage cannot be caused.
Description
Technical Field
The invention relates to the technical field of federal learning modeling, in particular to a multi-task unidirectional capsule network resource scheduling method based on federal learning.
Background
Federal learning is essentially a distributed machine learning technique, or machine learning framework. On the basis of ensuring data privacy safety and legal compliance, the method and the device realize common modeling and improve the effect of an AI model. In the aspect of application scenarios, the dimension information of a single sample is mainly increased by expanding dimension features, or the sample diversity is increased by increasing the sample size of the same dimension features, and the problem that the accuracy and robustness of a unilaterally trained machine learning model are insufficient is solved on the basis of ensuring that each party holds information.
In actual life, federal learning has wide application requirements, such as anti-fraud models, recommendation models for accurate marketing and the like, and the accuracy and the universality of the federal model trained by expanding the dimensional characteristics of sample data and sample size can be improved, so that individuals or enterprises can be helped to avoid risks.
The federal modeling training task is more complex and longer than the general distributed model and the single-machine model, which leads to the failure of the whole training directly once communication problems such as communication timeout occur. This may eventually render the model useless as the various parties cannot communicate with each other to coordinate the troubleshooting problem. In addition, in the existing distributed federal learning framework, the multi-task network communication block still adopts pressure-filling communication, namely, a task packet is sent under the condition of full bandwidth through multiple threads or multiple processes, the utilization efficiency of the bandwidth can be ensured by the mode, but inevitable problems exist, for example, when a larger model is trained, the data needing communication often reaches more than a plurality of GB (gigabytes), longer communication time is needed under the condition of general bandwidth, according to the pressure-filling communication mode, communication is carried out first, other small tasks are impossible when needing to apply for the bandwidth, or waiting for a long time is needed, timeout is often caused, in addition, if a smaller model is trained, a participant starts a larger task, the larger task started later in the aspect of communication extrudes the task bandwidth started earlier, so that the smaller task needs a very long time to complete the training, and the user experience and the stability of model training are greatly reduced. Therefore, the existing federal learning and privacy calculation communication framework usually depends on a communication control module of a communication protocol, so that most communication broadband is in a congestion state under the condition of multitask and multi-party, the execution speed of the whole task is reduced due to the communication congestion, and besides, due to channel congestion, a newly-initiated task cannot be started normally, so that the system is abnormal.
Disclosure of Invention
The invention aims to provide a multi-task one-way capsule network resource scheduling method based on federal learning, which can ensure that each task is in complete communication and avoid the problem that all tasks cannot be normally completed because network channels are mutually extruded by two-way communication.
The embodiment of the invention is realized by the following steps:
in a first aspect, an embodiment of the present application provides a method for scheduling a network resource based on a federal learning multitask unidirectional capsule, which includes the following steps:
s110: the initiator initializes the task and the data source, simultaneously virtualizes the single communication coordination module and synchronously informs the participants of starting response;
s120: when multiple tasks are parallel, a data packet to be sent of any task is decomposed into a plurality of task communication packets;
s130: putting a plurality of task communication packets into a waiting cabin of the capsule cabin according to an agent proportion table;
s140: when the waiting cabin is completely loaded or the waiting time is completed, coordinating all communication according to the single communication coordination module to judge whether to use a channel to send the capsule cabin;
s150: the capsule cabin starts multithreading and sends the multithreading to a participant, and the participant puts data in the capsule cabin into a specified task queue according to tasks and waits for the tasks to be acquired;
s160: when any task acquires the waiting data, the next stage of the task is entered to complete the task, and S120 to S160 are repeatedly executed until all tasks are completed.
In some embodiments of the present invention, the step of S130 includes:
and dynamically adjusting the seat proportion table according to the number of the tasks of the current task.
In some embodiments of the present invention, after the step of S110, the method further includes:
and after the task is successfully created, synchronizing task information to the single communication coordination module to perform task preparation.
In some embodiments of the present invention, after the step of S130, the method further includes:
acquiring a capsule cabin state, and reporting the capsule cabin state to a single communication coordination module;
and the single communication coordination module judges the current path according to the state of the capsule cabin and counts the current path into the scheduling.
In some embodiments of the present invention, the step of S140 includes:
when the current path is ready to be completed, a capsule compartment transfer instruction is sent.
In a second aspect, an embodiment of the present application provides an electronic device, which includes a memory for storing one or more programs; a processor. The program or programs, when executed by a processor, implement the method of any of the first aspects as described above.
In a third aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the method according to any one of the above first aspects.
Compared with the prior art, the embodiment of the invention has at least the following advantages or beneficial effects:
the invention provides a multi-task unidirectional capsule network resource scheduling method based on federal learning, which comprises the following steps: s110: the initiator initializes the task and the data source, simultaneously virtualizes the single communication coordination module, and synchronously informs the participants of starting response. S120: when multiple tasks are parallel, a data packet to be sent of any task is decomposed into a plurality of task communication packets. S130: and placing the plurality of task communication packets into a waiting cabin of the capsule cabin according to the seat proportion table. By means of the capsule cabin, all tasks of one node have a unified communication mode, all tasks can be guaranteed to be executed, all tasks can be considered due to the fact that the seat proportion table is dynamically adjusted according to the tasks of the current node, and communication coordination of a single communication coordination module is facilitated. S140: and when the waiting cabin is completely loaded or the waiting time is completed, coordinating all communication according to the single communication coordination module to judge whether to use the channel to send the capsule cabin. All communication is coordinated through the single communication coordination module, so that no jam between every two participants is ensured, only one party is allowed to send data to the other party at the same time, the condition of bidirectional data sending is avoided, and the concept of a capsule cabin is matched, so that the global routing is more convenient. S150: and the capsule cabin starts multithreading and sends the multithreading to the participators, and the participators put the data in the capsule cabin into a specified task queue according to the tasks and wait for the task to be acquired. Full bandwidth can be achieved using multiple threads while occupying as little computing resources as possible. The single communication coordination module can ensure that only one capsule cabin is used for transmission on each communication line, and the multithreading full-bandwidth transmission mode ensures that the communication line on each communication line is optimal. S160: and when any task acquires the waiting data, entering the next stage of the task to complete the task, and repeatedly executing S120-S160 until all tasks are completed. The method enables a plurality of tasks to be smoothly completed through a controllable communication mode, and the condition of local communication blockage is not caused.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a flowchart of a method for scheduling resources of a multi-task unidirectional capsule network based on federal learning according to an embodiment of the present invention;
fig. 2 is a flowchart of another method for scheduling resources of a unidirectional capsule network based on federal learning according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of an interface delivery capsule provided by an embodiment of the present invention;
fig. 4 is a flowchart of another method for scheduling resources in a multi-task unidirectional capsule network based on federal learning according to an embodiment of the present invention;
fig. 5 is a schematic structural block diagram of an electronic device according to an embodiment of the present invention.
An icon: 101-a memory; 102-a processor; 103-communication interface.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations.
Examples
Referring to fig. 1, fig. 1 is a flowchart illustrating a method for scheduling a multi-task unidirectional capsule network resource based on federal learning according to an embodiment of the present invention. The embodiment of the application provides a multi-task unidirectional capsule network resource scheduling method based on federal learning, which comprises the following steps:
s110: the initiator initializes the task and the data source, simultaneously virtualizes the single communication coordination module and synchronously informs the participants of starting response;
s120: when multiple tasks are parallel, a data packet to be sent of any task is decomposed into a plurality of task communication packets;
s130: putting a plurality of task communication packets into a waiting cabin of the capsule cabin according to an agent proportion table;
wherein, the seat proportion table is shown in table 1:
table 1: seat proportion table
In some embodiments of this embodiment, the step S130 includes: and dynamically adjusting the seat proportion table according to the number of the tasks of the current task. Specifically, the weight of each task can be dynamically adjusted by maintaining a seat proportion table of the capsule cabin, the weight corresponds to the seat of the capsule cabin, taking the capsule cabin with 200 seats as an example, the weight is 0.3, 60 seats are obtained, and 60 data packets can be sent at the same time. Therefore, each task can be ensured to have the self-executed bandwidth through the seat dynamic control of the capsule cabin.
In the implementation process, all tasks of one node have a unified communication mode in a capsule cabin mode, and all tasks can be executed.
S140: when the waiting cabin is completely loaded or the waiting time is completed, coordinating all communication according to the single communication coordination module to judge whether to use a channel to send the capsule cabin;
specifically, when the federal study is used for establishing a task, each participant can only know data information and cannot know respective communication conditions, and through the single communication control module, the communication of the participants who are black boxes can be smooth and efficient. All communication is coordinated through the single communication coordination module, so that no jam between every two participants is ensured, only one party is allowed to send data to the other party at the same time, the condition of bidirectional data sending is avoided, and the concept of a capsule cabin is matched, so that the global routing is more convenient.
S150: the capsule cabin starts multithreading and sends the multithreading to a participant, and the participant puts data in the capsule cabin into a specified task queue according to tasks and waits for the tasks to be acquired;
specifically, the capsule cabin transmission is transmitted in a multithreading mode and full bandwidth, and the single communication coordination module is informed after the transmission is finished. Full bandwidth can be achieved using multithreading while occupying as little computing resources as possible. The single communication coordination module can ensure that only one capsule cabin is used for transmission on each communication line, and the multithreading full-bandwidth transmission mode ensures that the communication line on each communication line is optimal.
S160: when any task acquires the waiting data, the next stage of the task is entered to complete the task, and S120 to S160 are repeatedly executed until all tasks are completed.
Specifically, the method enables a plurality of tasks to be smoothly completed through a controllable communication mode, and the condition of local communication blockage is not caused.
In the implementation process, the method is used for carrying out upgrading optimization on the aspect of multi-task situation network communication in the federal modeling process, ensuring that the model can carry out stable and safe training and prediction in the multi-task federal modeling and prediction process, sending and receiving encrypted data packets in the training and prediction process, and ensuring the tradition of network communication and the proportion and weight of each model in the aspect of communication. The method combines various technologies of bandwidth control, capsule cabin, communication coordination and automatic optimization, and when multiple tasks are performed in parallel, the problem that smaller training tasks started later cannot be started or overtime communication is caused by the current larger model training task is solved while the training is started first and then started. Meanwhile, a plurality of participants simultaneously initiate tasks to perform multi-task coordination and share the network channel, complete communication of each task is guaranteed in a capsule cabin mode, and the problem that all tasks cannot be normally completed due to mutual extrusion of the network channel through bidirectional communication is avoided.
Referring to fig. 2 and fig. 3, fig. 2 is a flowchart illustrating another implementation of a resource scheduling method for a multi-task unidirectional capsule network based on federal learning according to an embodiment of the present invention, and fig. 3 is a schematic diagram illustrating an interface sending capsule according to an embodiment of the present invention. Firstly, fang Jia is initiated to virtualize a single item communication coordination node, and after a new task is created, task preparation is carried out on synchronous task information of a single item communication coordination module. In the process of initiating Fang Jia communication with a participant B, when a task A, a task B and a task C are parallel, for any task of the task A, the task B or the task C, subpackaging the communication content of any task to obtain a plurality of task communication packets, and transferring the plurality of task communication packets into a waiting cabin of a capsule cabin according to a seat proportion table of the capsule cabin. When the waiting cabin is completely loaded or the waiting time is completed, all communication is coordinated according to the single communication coordination module, the capsule cabin is sent to the participant B in a multithreading mode, the task A is placed in the task A queue, the task B is placed in the task B queue, and the task C is placed in the task C queue.
In addition, the communications between the participant Fang Yi and the participant c, and the communications between the initiator Fang Jia and the participant c are the same as the communications between the initiator Fang Jia and the participant b, and therefore the description is not repeated here.
In some embodiments of this embodiment, after the step of S110, the method further includes:
and after the task is successfully created, synchronizing task information to the single communication coordination module to perform task preparation.
Referring to fig. 4, fig. 4 is a flowchart illustrating a resource scheduling method for a multi-task unidirectional capsule network based on federal learning according to another embodiment of the present invention. In some embodiments of this embodiment, after the step of S130, the method further includes:
acquiring a capsule cabin state, and reporting the capsule cabin state to a single communication coordination module;
and the single communication coordination module judges the current path according to the state of the capsule cabin and counts the current path into the scheduling.
Specifically, the method reports the capsule cabin state to a single communication coordination module, the single communication coordination module judges the current path and counts the current path into a scheduling period, and the capsule cabin is transmitted after the path is available.
In some embodiments of this embodiment, the step S140 includes:
when the current path is ready to be completed, a capsule compartment transfer instruction is sent.
Referring to fig. 5, fig. 5 is a schematic structural block diagram of an electronic device according to an embodiment of the present disclosure. The electronic device comprises a memory 101, a processor 102 and a communication interface 103, wherein the memory 101, the processor 102 and the communication interface 103 are electrically connected to each other directly or indirectly to realize data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The memory 101 may be used to store software programs and modules, and the processor 102 executes the software programs and modules stored in the memory 101 to thereby execute various functional applications and data processing. The communication interface 103 may be used for communicating signaling or data with other node devices.
The Memory 101 may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Read-Only Memory (EPROM), an electrically Erasable Read-Only Memory (EEPROM), and the like.
The processor 102 may be an integrated circuit chip having signal processing capabilities. The Processor 102 may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components.
It will be appreciated that the configuration shown in fig. 5 is merely illustrative and that the electronic device may include more or fewer components than shown in fig. 5 or have a different configuration than shown in fig. 5. The components shown in fig. 5 may be implemented in hardware, software, or a combination thereof.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The apparatus embodiments described above are merely illustrative, and for example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist alone, or two or more modules may be integrated to form an independent part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
It will be evident to those skilled in the art that the present application is not limited to the details of the foregoing illustrative embodiments, and that the present application may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Claims (7)
1. A multi-task unidirectional capsule network resource scheduling method based on federal learning is characterized by comprising the following steps:
s110: the initiator initializes the task and the data source, simultaneously virtualizes the single communication coordination module and synchronously informs the participants of starting response;
s120: when multiple tasks are parallel, a data packet to be sent of any task is decomposed into a plurality of task communication packets;
s130: putting a plurality of task communication packets into a waiting cabin of the capsule cabin according to an agent proportion table;
s140: when the waiting cabin is completely loaded or the waiting time is completed, coordinating all communication according to the single communication coordination module to judge whether a channel is used for sending the capsule cabin;
s150: the capsule cabin starts multithreading and sends the multithreading to a participant, and the participant puts data in the capsule cabin into a specified task queue according to tasks and waits for the tasks to be obtained;
s160: and when any task acquires the waiting data, entering the next stage of the task to complete the task, and repeatedly executing S120-S160 until all tasks are completed.
2. The method according to claim 1, wherein the step S130 comprises:
and dynamically adjusting the seat proportion table according to the number of tasks of the current task.
3. The method for scheduling resource of a federated learning-based multitask one-way capsule network as claimed in claim 1, wherein after the step of S110, further comprising:
and after the task is successfully created, synchronizing task information to the single communication coordination module to perform task preparation.
4. The method for scheduling resources for a multi-tasking unidirectional capsule network based on federal learning as claimed in claim 1, wherein the step S130 is followed by further comprising:
acquiring a capsule cabin state, and reporting the capsule cabin state to the single communication coordination module;
and the single communication coordination module judges the current path according to the capsule cabin state and counts the current path into the schedule.
5. The method according to claim 4, wherein the step S140 comprises:
when the current path is ready to be completed, a capsule compartment delivery instruction is sent.
6. An electronic device, comprising:
a memory for storing one or more programs;
a processor;
the one or more programs, when executed by the processor, implement the method of any of claims 1-5.
7. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-5.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211438714.1A CN115827224A (en) | 2022-11-17 | 2022-11-17 | Multi-task one-way capsule network resource scheduling method based on federal learning |
| PCT/CN2022/137216 WO2024103457A1 (en) | 2022-11-17 | 2022-12-07 | Multi-task unidirectional capsule network resource scheduling method based on federated learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211438714.1A CN115827224A (en) | 2022-11-17 | 2022-11-17 | Multi-task one-way capsule network resource scheduling method based on federal learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115827224A true CN115827224A (en) | 2023-03-21 |
Family
ID=85528705
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211438714.1A Pending CN115827224A (en) | 2022-11-17 | 2022-11-17 | Multi-task one-way capsule network resource scheduling method based on federal learning |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN115827224A (en) |
| WO (1) | WO2024103457A1 (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11010902B2 (en) * | 2018-06-04 | 2021-05-18 | University Of Central Florida Research Foundation, Inc. | Capsules for image analysis |
| CN111768008B (en) * | 2020-06-30 | 2023-06-16 | 平安科技（深圳）有限公司 | Federal learning method, apparatus, device, and storage medium |
| EP3992861A1 (en) * | 2020-07-17 | 2022-05-04 | Tata Consultancy Services Limited | System and method for parameter compression of capsule networks using deep features |
| CN113570069A (en) * | 2021-07-28 | 2021-10-29 | 神谱科技(上海)有限公司 | Model evaluation method for self-adaptive starting model training based on safe federal learning |
| CN114328432A (en) * | 2021-12-02 | 2022-04-12 | 京信数据科技有限公司 | Big data federal learning processing method and system |
-
2022
- 2022-11-17 CN CN202211438714.1A patent/CN115827224A/en active Pending
- 2022-12-07 WO PCT/CN2022/137216 patent/WO2024103457A1/en unknown
Also Published As
| Publication number | Publication date |
|---|---|
| WO2024103457A1 (en) | 2024-05-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10891177B2 (en) | Message management method and device, and storage medium | |
| CN107087019B (en) | Task scheduling method and device based on end cloud cooperative computing architecture | |
| US10599488B2 (en) | Multi-purpose events for notification and sequence control in multi-core processor systems | |
| CN111950988B (en) | Distributed workflow scheduling method and device, storage medium and electronic equipment | |
| US11321139B2 (en) | Streaming traffic pattern for public cloud auto scaling | |
| CN112449750A (en) | Log data collection method, log data collection device, storage medium, and log data collection system | |
| WO2019104713A1 (en) | Machine learning method, master node, work node, and system | |
| WO2017185615A1 (en) | Method for determining service status of service processing device and scheduling device | |
| US20220253268A1 (en) | Smart screen share reception indicator in a conference | |
| CN112256523A (en) | Service data processing method and device | |
| CN111586140A (en) | Data interaction method and server | |
| CN110727507B (en) | Message processing method and device, computer equipment and storage medium | |
| CN113342503B (en) | Real-time progress feedback method, device, equipment and storage medium | |
| CN115562887A (en) | Inter-core data communication method, system, device and medium based on data package | |
| CN112000446A (en) | Data transmission method and robot | |
| CN107294911B (en) | Data packet monitoring method and device, remote procedure call system and equipment | |
| CN113946389A (en) | Federal learning process execution optimization method, device, storage medium, and program product | |
| CN114138506A (en) | Message queue scheduling method and device, equipment, medium and product thereof | |
| CN115827224A (en) | Multi-task one-way capsule network resource scheduling method based on federal learning | |
| CN113556387A (en) | Edge gateway control method, system, device, electronic equipment and storage medium | |
| CN111124708B (en) | Microservice-oriented batch reasoning method, server and computer readable storage medium | |
| CN103761106A (en) | Process control method and process engine | |
| US20210149709A1 (en) | Method and apparatus for processing transaction | |
| CN111028329B (en) | Rendering graph providing method, device and equipment and storage medium | |
| CN115361382B (en) | Data processing method, device, equipment and storage medium based on data group |
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 |