KR20210121914A - System and method for operating artificial intelligence and machine learning model - Google Patents

Abstract

The present invention relates to a method for operating an artificial intelligence and machine learning (AI/ML) model in a federated learning (FL) client including an AI/ML service process unit, a local AI/ML model storage and an FL process unit. An embodiment of the present invention comprises the steps of: the AI/ML service process unit requesting the download of an AI/ML model from the AI/ML model storage; storing the AI/ML model in the local AI/ML model storage after the AI/ML model is downloaded in response to the request of download; the AI/ML service process unit reading and storing the AI/ML model from the local AI/ML model storage to provide a service; the FL process unit determining whether or not performance conditions for federated learning are satisfied; reading and storing the AI/ML model from the local AI/ML model storage to perform federated learning as the performance conditions for FL are satisfied; and providing, to an FL server, the AI/ML model renewed as the performance of the federated learning is finished. Therefore, provided are a system and a method for operating an AI/ML model wherein an AI/ML service and an FL service can be operated efficiently.

Description

SYSTEM AND METHOD FOR OPERATING ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING MODEL

The present invention relates to a system and method for operating artificial intelligence and machine learning models, and more particularly, to a system and method for effectively operating an AI/ML model between a federated learning processor and other processors.

Recently, as cloud technology, artificial intelligence, and machine learning (hereinafter referred to as AI/ML) technology develop, various AI/ML services are being provided with high accuracy.

In order to maintain such a high degree of accuracy, a large amount of data for learning and high-performance computing resources capable of processing it are required.

To this end, in the prior art, learning data was collected in a high-performance server or cloud, etc. and learning was carried out. However, as issues such as personal information protection have recently been raised, the problem of storing large amounts of learning data in the cloud has been raised.

Federated Learning (FL) does not store data in one place to solve this problem, but stores it only in the data generator, then proceeds with learning in the data generator and collects only the learning results on an independent server to create AI/ML It was presented as a way to improve the performance of the model.

This federated learning has the effect of fundamentally blocking the leakage of personal information because the data is not gathered in one place and the learning itself proceeds from the data generator.

However, federated learning is still being implemented at the level of simulation through computing resources on the cloud, and there is a problem in that it is difficult to efficiently use computing resources when implementing because no specific method is presented for connection with services.

An embodiment of the present invention is artificial intelligence and machine learning that minimizes interference with the existing AI/ML service in operating the federated learning between the real server and the client so that the AI/ML service and the federated learning service can efficiently interoperate. A model operating system and method are provided.

However, the technical task to be achieved by the present embodiment is not limited to the technical task as described above, and other technical tasks may exist.

As a technical means for achieving the above technical task, an artificial intelligence and machine learning (AI/ML) service process unit, a local AI/ML model repository, a Federated Learning (FL) process according to the first aspect of the present invention A method of operating an AI/ML model in an FL client including a unit includes: requesting, by the AI/ML service processing unit, to download an AI/ML model from an AI/ML model repository; storing the AI/ML model in the local AI/ML model storage as the AI/ML model is downloaded in response to the download request; providing, by the AI/ML service processing unit, reading and storing the AI/ML model from the local AI/ML model storage and providing a service; determining, by the FL processing unit, whether a federated learning performance condition is satisfied; performing federated learning by reading and storing the AI/ML model from the local AI/ML model storage as the FL performance condition is satisfied; and providing the updated AI/ML model to the FL server as the federated learning is completed.

In some embodiments of the present invention, the local AI/ML model storage may be stored in a shareable volume between an AI/ML service processing unit in the FL client and a FL processing unit.

In some embodiments of the present invention, when the AI/ML service processing unit and the FL processing unit are configured as a Docker container, the local AI/ML model storage may be configured as a shared volume of a host.

In some embodiments of the present invention, the step of determining whether the FL processing unit satisfies the federated learning performance condition includes whether the FL client is powered on, whether it is idle for a preset time, and updating the AI/ML model It may be determined based on at least one of whether or not the collected data is satisfied.

In some embodiments of the present invention, the FL server may collect the updated AI/ML model from at least one FL client, and update it to an integrated model based on the collected AI/ML model when an update condition is satisfied.

In some embodiments of the present invention, the FL server may register the updated unified model in the AI/ML model repository as the update of the unified model is completed.

In some embodiments of the present invention, the method may further include the step of checking, by the AI/ML service processing unit, whether the integrated model of the AI/ML model storage is updated every preset period. In this case, the step of the AI/ML service processing unit requesting the download of the AI/ML model from the AI/ML model storage includes: When it is confirmed that the integrated model is updated, the You can request a download.

In some embodiments of the present invention, the step of checking, by the AI/ML service processing unit, whether to update the integrated model of the AI/ML model storage every preset period, includes the version information of the integrated model and the FL processing unit. Whether the update is made may be checked based on one or more of the updated version information of the AI/ML model.

In addition, the artificial intelligence and machine learning (AI/ML) model operating system according to the second aspect of the present invention is an AI/ML service processing unit that receives an AI/ML model and provides an AI/ML service, the AI/ML model FL that updates the AI/ML model by performing Federated Learning (FL) by reading and storing the AI/ML model from the local AI/ML model storage and the local AI/ML model storage that stores and provides at least one FL client including a processing unit; an AI/ML model storage that provides the AI/ML model to the local AI/ML model storage according to a request of the AI/ML service processing unit; and a FL server that collects the updated AI/ML model, updates the collected AI/ML model as an integrated model as an update condition is satisfied, and registers it in the AI/ML model repository.

In some embodiments of the present invention, the local AI/ML model storage may be stored in a shareable volume between an AI/ML service processing unit in the FL client and a FL processing unit.

In some embodiments of the present invention, the AI/ML service processing unit checks whether the integrated model of the AI/ML model storage is updated at every preset period, and when the integrated model is updated as a result of the check, the information on the integrated model is updated. You can request a download.

In addition, the Federated Learning (FL) client for artificial intelligence and machine learning (AI/ML) model operation according to the third aspect of the present invention is based on the AI/ML model provided from the AI/ML model repository. AI/ML service processing unit that provides ML service, local AI that receives the AI/ML model requested from the AI/ML service processing unit, stores it, and provides it to the AI/ML service processing unit Based on whether the /ML model storage and federated learning performance conditions are satisfied, federated learning is performed by reading and storing the AI/ML model from the local AI/ML model storage, and the updated AI/ML model is transferred to the FL server FL processing unit provided by

In addition to this, another method for implementing the present invention, another system, and a computer-readable recording medium for recording a computer program for executing the method may be further provided.

Other specific details of the invention are included in the detailed description and drawings.

According to any one of the above-described problem solving means of the present invention, in the joint learning operation between the FL server and the FL client, communication efficiency is achieved by reducing the data bandwidth for transmission of AI/ML models between the server and the client. can do.

In addition, by minimizing the interference with the existing AI/ML service in the operation of federated learning, the AI/ML service and the FL service can be efficiently interoperated.

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a diagram for explaining a relationship between an AI/ML process and a federated learning process in the prior art.
2 is a diagram for explaining an AI/ML model operating system according to an embodiment of the present invention.
3 is a block diagram of an AI/ML model operating system according to an embodiment of the present invention.
4 is a flowchart of an AI/ML model operating method according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement them. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted.

In the entire specification, when a part "includes" a certain element, it means that other elements may be further included, rather than excluding other elements, unless otherwise stated.

1 is a diagram for explaining the relationship between the AI/ML service process 11 and the federated learning process 12 in the prior art.

Federated Learning (FL) is a learning technique proposed to prevent problems such as leakage of personal information that occurs in the existing large data learning process and to efficiently utilize computing resources.

The federated learning is operated in an environment in which the FL server 20 and the FL client 10 exist, and is performed according to the following process.

First, the FL server 20 transmits to the FL client 10 a learning technique to be used in the FL client 10 together with the AI/ML model generated primarily.

The FL client 10 collects local data at the same time as receiving the corresponding AI/ML model and learning technique, and if the amount of collected data is sufficient for training, it is applied to the learning technique together with the AI/ML model that has received the data. to update the AI/ML model.

Then, when the update of the AI/ML model is completed, the FL client 10 transmits the updated AI/ML model to the FL server 20 again, and the FL server 20 receives from the plurality of FL clients 10 . An integrated model is generated by aggregating the updated AI/ML models.

The integrated model generated in this way is again distributed to the FL client 10, and the above process is repeated.

This federated learning is a model update method through distributed learning, and has the advantage of protecting personal information, improving model performance, and making optimal use of computing resources. However, the implementation of federated learning has not yet been confirmed at the level of simulation using actual computing resources.

In addition, since federated learning exists in parallel with the actual AI/ML service, conflicts between processes may occur if the flow for connection and access to the existing AI/ML service operation is not defined when operating federated learning.

For example, when the updated AI/ML model is redistributed from the FL server 20 while the FL client 10 is providing an analysis service based on the AI/ML model distributed from the FL server 20, There may be problems with the stability of the AI/ML service in operation.

In addition, when the two processes 11 and 12 are independently operated, the AI/ML model is redundantly present in the FL client 10 . In other words, when each FL client 10 operates the AI/ML service and the federated learning service at the same time, as each is composed of an independent service or process 11, 12, there is a problem in the efficiency of AI/ML model management. do. For example, when the process unit 11 for AI/ML service holds an AI/ML model for operation of the corresponding service, and the process unit 12 for FL service also operates an independent AI/ML model , a single AI/ML model is redundantly stored in the corresponding FL client 10, and network and storage resources are wasted.

In addition, a separate process is required to distribute the updated integrated model again after the update of the integrated model is performed in the FL server 20 . In addition, an unnecessary network load is required in the update operation of each model, and there is a possibility that a conflict in resource utilization between services may occur.

In order to solve this problem, an embodiment of the present invention provides a local AI/ML model storage 120 in the FL client 100, and at the same time, a separate space between the FL client 100 and the FL server 300 is provided. Through the structure provided with the AI/ML model storage 200, AI/ML-based services and federated learning-based services can be mutually independent and organically linked.

Hereinafter, the AI/ML model operating system 1 according to an embodiment of the present invention will be described with reference to FIGS. 2 to 3 .

2 is a diagram for explaining the AI/ML model operating system 1 according to an embodiment of the present invention. 3 is a block diagram of an AI/ML model operating system 1 according to an embodiment of the present invention.

Referring to FIG. 2 , the AI/ML model operating system 1 according to an embodiment of the present invention includes at least one FL client 100 , an AI/ML model storage 200 , and a FL server 300 . include

The FL client 100 includes an AI/ML service processing unit 110 , a local AI/ML model store 120 , and a FL processing unit 130 .

The AI/ML service processing unit 110 provides AI/ML services based on the AI/ML model provided from the AI/ML model storage 200 .

The local AI / ML model storage 120 receives the AI / ML model requested from the AI / ML service processing unit 110 from the AI / ML model storage 200 and stores it, and the AI / ML service processing unit 110 . provides this with

In an embodiment, the local AI/ML model storage 120 may be stored in a shareable volume between the AI/ML service processing unit 110 and the FL processing unit 130 in the FL client 100 . For example, when each of the process units 110 and 130 is configured as a Docker container, the local AI/ML model storage 120 may be configured as a shared volume of the host.

The FL processing unit 130 determines whether or not the federated learning execution condition is satisfied, and when the execution condition is satisfied, the AI/ML model is read from the local AI/ML model storage 120 and stored, and then federated learning is performed. . And the AI/ML model updated according to the result of performing the federated learning is provided to the FL server 300 .

The AI/ML model storage 200 provides the AI/ML model to the local AI/ML model storage 120 according to the request of the AI/ML service processing unit 110 . When the AI/ML model is stored in the local AI/ML model storage 120 according to such a request, the AI/ML service processing unit 110 reads the AI/ML model, stores it in memory, and provides AI/ML service. do.

The FL server 300 collects the updated AI/ML model from the FL clients 100, updates the collected AI/ML model as an integrated model as the update condition is satisfied, and stores it in the AI/ML model storage 200. register

Accordingly, the integrated model is updated with a new version of the AI/ML model in the AI/ML model repository 200 .

In addition, the AI/ML service process unit 110 checks whether the integrated model of the AI/ML model storage 200 is updated every preset cycle, and when the integrated model is updated as a result of the check, the AI/ML service process unit 110 requests a download for the integrated model. A new version of the AI/ML model stored in the ML model storage 200 is provided.

On the other hand, in an embodiment of the present invention, the FL client 100, the AI/ML model storage 200 and the FL server 300 are the communication module 410, the memory 420 and the memory 420 as shown in FIG. It may be configured to include a processor 430 that executes a program stored in the .

The communication module 410 is preferably configured as a wireless communication module, but this does not exclude a wired communication module. The wireless communication module may be implemented by wireless LAN (WLAN), Bluetooth, HDR WPAN, UWB, ZigBee, Impulse Radio, 60GHz WPAN, Binary-CDMA, wireless USB technology, wireless HDMI technology, and the like. In addition, the wired communication module may be implemented as a power line communication device, a telephone line communication device, a cable home (MoCA), Ethernet, IEEE1294, an integrated wired home network, and an RS-485 control device.

The memory 420 collectively refers to a non-volatile storage device and a volatile storage device that continuously maintain stored information even when power is not supplied. For example, the memory 420 may include a compact flash (CF) card, a secure digital (SD) card, a memory stick, a solid-state drive (SSD), and a micro SD card. NAND flash memory such as cards, magnetic computer storage devices such as hard disk drives (HDDs), etc., and optical disc drives such as CD-ROMs and DVD-ROMs. can

For reference, the components described in the embodiment of the present invention may be implemented in the form of software or hardware such as a field programmable gate array (FPGA) or an application specific integrated circuit (ASIC), and may perform predetermined roles.

However, 'components' are not limited to software or hardware, and each component may be configured to reside in an addressable storage medium or to reproduce one or more processors.

Thus, as an example, a component includes components such as software components, object-oriented software components, class components and task components, and processes, functions, properties, procedures, sub It includes routines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables.

Components and functions provided within the components may be combined into a smaller number of components or further divided into additional components.

Hereinafter, an AI/ML model operating method in the FL client 100 will be described with reference to FIG. 4 .

4 is a flowchart of an AI/ML model operating method according to an embodiment of the present invention.

At this time, in the AI/ML model operating method according to an embodiment of the present invention, the AI/ML service processing unit 110 for providing the AI service in the FL client 100 and the FL processing unit for updating the AI/ML model (130) is in operation.

In addition, the initial AI/ML model is stored in a separate AI/ML model storage 200, and AI/ML models are not yet downloaded or loaded into memory in each of the processing units 110 and 130 in the FL client 100. Assume that it is not

First, the AI/ML service processing unit 110 requests the download of the initial AI/ML model from the AI/ML model storage 200 (S110).

Next, in response to the download request, the AI/ML model is downloaded and stored from the AI/ML model storage 200 to the local AI/ML model storage 120 ( S120 ).

Next, the AI/ML service processing unit 110 reads the AI/ML model from the local AI/ML model storage 120 and stores it in a memory, and then provides the AI/ML service (S130).

Thereafter, the FL processing unit 130 periodically determines whether or not the federated learning execution condition is satisfied ( S140 ), and as a result of the determination, the AI/ML from the local AI/ML model storage 120 satisfies the federated learning execution condition. After reading the model and storing it in the memory, federated learning is performed (S150). In this process, in an embodiment of the present invention, the AI/ML service processing unit 110 and the FL processing unit 130 share the same stored AI/ML model.

In one embodiment, the federated learning performance condition is whether the power of the FL client 100 is supplied, whether it is idle for a preset time, and whether the amount of collected data for updating the AI/ML model is satisfied One or more of them may be applied in combination.

Next, as the federated learning is completed, the updated AI/ML model is provided to the FL server 300 (S160).

Meanwhile, the FL server 300 collects updated AI/ML models from at least one FL client 100 . Then, it is determined whether the update condition is satisfied, and when the update condition is satisfied, the integrated model is updated based on the collected AI/ML model.

When the update to the integrated model is completed, the FL server 300 registers the updated integrated model in the AI/ML model storage 200 .

Accordingly, the AI/ML service process unit 110 checks whether the AI/ML model storage 200 is updated to the integrated model at every preset cycle, and when the integrated model is updated as a result of the check, the AI/ML model storage The AI/ML model stored in advance is updated by requesting the download of the integrated model from (200).

At this time, according to an embodiment of the present invention, the AI/ML model should include additional information that can confirm each update information throughout the entire process, and version information for the integrated model as such additional information, the FL process unit 130 It must contain at least one of the version information of the AI/ML model updated by .

Meanwhile, in the above description, steps S110 to S160 may be further divided into additional steps or combined into fewer steps according to an embodiment of the present invention. In addition, some steps may be omitted as necessary, and the order between steps may be changed. In addition, even if other contents are omitted, the contents of FIGS. 2 to 3 may be applied to the AI/ML model operating method of FIG. 4 .

An embodiment of the present invention may also be implemented in the form of a computer program stored in a medium executed by a computer or a recording medium including instructions executable by the computer. Computer-readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. Also, computer-readable media may include both computer storage media and communication media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Communication media typically includes computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave, or other transport mechanism, and includes any information delivery media.

Although the methods and systems of the present invention have been described with reference to specific embodiments, some or all of their components or operations may be implemented using a computer system having a general purpose hardware architecture.

The above description of the present invention is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a dispersed form, and likewise components described as distributed may be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

1: AI/ML model operating system
100: FL client
110: AI/ML service processing unit
120: Local AI/ML Model Repository
130: FL process unit
200: AI/ML Model Repository
300: FL Server
410: communication module,
420: memory
430: processor

Claims (12)

In an AI/ML model operation method in an FL client including an artificial intelligence and machine learning (AI/ML) service process unit, a local AI/ML model repository, and a federated learning (FL) process unit,
requesting, by the AI/ML service processing unit, to download the AI/ML model from the AI/ML model repository;
storing the AI/ML model in the local AI/ML model storage as the AI/ML model is downloaded in response to the download request;
providing, by the AI/ML service processing unit, reading and storing the AI/ML model from the local AI/ML model storage and providing a service;
determining, by the FL processing unit, whether a federated learning performance condition is satisfied;
performing federated learning by reading and storing the AI/ML model from the local AI/ML model storage as the FL performance condition is satisfied; and
AI/ML model operating method comprising the step of providing the updated AI/ML model to the FL server as the federated learning is completed.
According to claim 1,
wherein the local AI/ML model storage is stored in a shareable volume between the AI/ML service processing unit in the FL client and the FL processing unit.
3. The method of claim 2,
When the AI/ML service processing unit and the FL processing unit are configured as a Docker container, the local AI/ML model storage is configured as a shared volume of a host.
According to claim 1,
The step of determining whether the FL processing unit satisfies the federated learning performance condition includes:
The AI/ML model operating method is determined based on at least one of whether the FL client is powered on, whether it is idle for a preset time, and whether collection data for updating the AI/ML model is satisfied.
According to claim 1,
The FL server collects the updated AI/ML model from at least one FL client, and when the update condition is satisfied, the AI/ML model operating method is updated to an integrated model based on the collected AI/ML model.
6. The method of claim 5,
The FL server is an AI/ML model operating method that registers the updated integrated model in the AI/ML model storage as the update of the integrated model is completed.
6. The method of claim 5,
The method further comprising the step of confirming, by the AI/ML service processing unit, whether the integrated model of the AI/ML model storage is updated every preset period,
The step of the AI/ML service processing unit requesting the download of the AI/ML model from the AI/ML model storage comprises:
When it is confirmed that the integrated model has been updated, the AI/ML model operating method of requesting the download of the integrated model from the AI/ML model repository.
8. The method of claim 7,
The step of the AI/ML service processing unit confirming whether to update the integrated model of the AI/ML model storage every preset period comprises:
The AI/ML model operating method of confirming whether the update is made based on at least one of the version information of the integrated model and the version information of the AI/ML model updated by the FL processing unit.
In the artificial intelligence and machine learning (AI/ML) model operating system,
An AI/ML service processing unit that receives an AI/ML model and provides an AI/ML service, a local AI/ML model repository that stores and provides the AI/ML model, and the AI/ML from the local AI/ML model repository At least one FL client including a FL processing unit that reads and stores the model to perform Federated Learning (FL) to update the AI/ML model;
an AI/ML model storage that provides the AI/ML model to the local AI/ML model storage according to the request of the AI/ML service processing unit;
AI/ML comprising a FL server that collects the updated AI/ML model from the FL client, updates the collected AI/ML model as an integrated model as an update condition is satisfied, and registers it in the AI/ML model repository model operating system.
10. The method of claim 9,
and the local AI/ML model storage is stored in a shareable volume between the AI/ML service processing unit and the FL processing unit in the FL client.
10. The method of claim 9,
The AI/ML service processing unit checks whether the integrated model of the AI/ML model storage is updated every preset period, and requests a download for the integrated model when the integrated model is updated as a result of the check. ML model operating system.
In a Federated Learning (FL) client for artificial intelligence and machine learning (AI/ML) model operation,
an AI/ML service processing unit that provides AI/ML services based on AI/ML models provided from the AI/ML model repository;
A local AI/ML model storage that receives and stores the AI/ML model requested from the AI/ML service processing unit from the AI/ML model storage and provides it to the AI/ML service processing unit;
FL process for performing federated learning by reading and storing the AI/ML model from the local AI/ML model storage based on whether the federated learning performance condition is satisfied, and providing the updated AI/ML model to the FL server FL clients with units.

Images