A communication-optimized federated learning method based on self-guided evolutionary strategy

Abstract

The present invention discloses a communication-optimized federated learning method based on a self-guided evolutionary strategy, belonging to the field of privacy computing technology. The method comprises the following steps: a user uses the global model gradient vectors from previous rounds to divide the real space into a main space and its orthogonal complement space, and extracts pseudo-random evaluation vectors from each of the two spaces; the high-dimensional model gradient vector is converted into fitness values ​​of multiple pseudo-random evaluation vectors and sent to a server; the server aggregates local fitness values ​​and uses the evaluation vectors to calculate a global gradient vector, and simultaneously sends the global fitness value to the user. The user and server calculate the global gradient vector for the current round based on the global fitness value and the pseudo-random evaluation vector, and update the global model. The present invention not only significantly reduces communication overhead, but also improves the effectiveness of the search direction by adaptively utilizing historical estimated gradients, thereby accelerating convergence and improving model performance.

Description

Technical Field

[0001] The present invention belongs to the field of privacy computing technology, and in particular relates to a communication optimization federated learning method based on a self-guided evolutionary strategy. Background Art

[0002] With the development of big data and cloud computing technologies, machine learning plays an increasingly important role in fields such as autonomous driving, speech recognition, image classification, and disease detection. In order to better extract valuable information from massive amounts of data, machine learning systems are often deployed in architectures containing thousands of processors. However, traditional centralized learning methods face challenges in data privacy and security, especially in the healthcare and financial industries. Federated Learning (FL), as a distributed learning framework, allows multiple devices (called users) to jointly train a model without sharing original data, aiming to address the above problems. However, one of the main bottlenecks of FL is inefficient communication, especially when dealing with deep learning models with a large number of parameters. Therefore, studying how to reduce communication costs has become an important topic in the field of federated learning.

[0003] To alleviate the communication bottleneck in federated learning, researchers have proposed a variety of techniques. These primarily include model compression, knowledge distillation, and client-side update sampling. While existing techniques for optimizing communication in federated learning offer their own strengths, they also present numerous limitations and challenges. Existing methods often require complex preprocessing steps, compromise model accuracy and robustness, or prove difficult to implement in large-scale distributed environments. Therefore, it is crucial to find new approaches that significantly reduce communication requirements while maintaining good model performance. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention proposes a communication optimization federated learning method based on a self-guided evolutionary strategy to solve the problems existing in the above-mentioned prior art.

[0005] To achieve the above objectives, the present invention provides a communication optimization federated learning method based on a self-guided evolutionary strategy, comprising:

[0006] Step 1: The server initializes the global model parameters and random number seed and sends them to each user;

[0007] Step 2: The user trains with local data to obtain a local gradient vector; a subspace is divided from the real number space based on the global gradient vector of a preset number of rounds and the random number seed, and a number of pseudo-random evaluation vectors are extracted from the subspace; the local fitness value of each pseudo-random vector with respect to the local gradient vector is calculated and fed back to the server;

[0008] Step 3: The server calculates the global fitness value based on the local fitness value fed back by each user and sends it to each user;

[0009] Step 4: The user and the server calculate the global gradient vector of this round based on the global fitness value and the pseudo-random evaluation vector, and update the global model;

[0010] Repeat steps 2 to 4 to continuously update the global model.

[0011] Optionally, the process of obtaining the pseudo-random evaluation vector in step 2 includes:

[0012] The user trains the local model using local data to obtain a local gradient vector. If the number of training rounds for the local model is less than or equal to a preset number of rounds, a pseudo-random gradient evaluation vector is generated based on the random number seed. If the number of training rounds for the local model is greater than the preset number of rounds, a subspace is divided from the real number space based on the global gradient vector and the random number seed, and several pseudo-random evaluation vectors are extracted from the subspace.

[0013] Optionally, in step 2, the pseudo-random gradient evaluation vector generated based on the random number seed obeys a normal distribution.

[0014] Optionally, in step 2, dividing a subspace from the real number space based on the global gradient vector and the random number seed, and extracting a plurality of pseudo-random evaluation vectors from the subspace includes:

[0015] The matrix composed of the global gradient vectors of the first few rounds of a preset number of rounds is subjected to matrix decomposition and divided into two parts: the main space and the orthogonal complement space. Sampling is performed in the main space and the orthogonal complement space according to a preset sampling ratio to obtain a number of pseudo-random evaluation vectors.

[0016] Optionally, in step 2, the formula for calculating the local fitness value of each pseudo-random vector to the local gradient vector is as follows:

[0017]

[0018] in, represents the local gradient vector obtained by user u in the tth round of training, represents the local fitness value of the i-th pseudo-random gradient vector, and m is the number of pseudo-random gradient vectors.

[0019] Optionally, in step 3, the formula for calculating the global fitness value is as follows:

[0020]

[0021] Among them, w uis the weight of user u, U is the total number of users, f i t is the global fitness value.

[0022] Optionally, in step four, the server generates a pseudo-random evaluation vector set using the method of step two, the server and the user respectively calculate the global gradient of this round and update the model parameters, and the user saves the global gradient for the next round of training.

[0023] Optionally, the formula for calculating the global gradient and model parameters of this round is as follows:

[0024]

[0025] θ t+1 =θ t -η*g t

[0026] Among them, g t is the global gradient obtained by aggregation in the tth round, η is the learning rate, θ t+1 is the t+1th round global model, m is the number of pseudo-random gradient evaluation vectors, is the pseudorandom gradient evaluation vector.

[0027] Compared with the prior art, the present invention has the following advantages and technical effects:

[0028] Users replace local gradient updates with fitness values ​​and upload them to the server. The server then replaces global gradients with global fitness values ​​and sends them to each user. Compared to high-dimensional gradient vectors, transmitting fitness values ​​significantly reduces communication overhead. Furthermore, because the pseudo-random gradient evaluation vectors are generated under the guidance of global gradients from previous rounds, they possess a certain self-bootstrapping property, resulting in a higher quality pseudo-random gradient evaluation vector set. This significantly improves the convergence efficiency of federated learning compared to traditional evolutionary strategies. This method not only significantly reduces communication overhead but also enhances the effectiveness of search directions by adaptively utilizing historical estimated gradients, thereby accelerating convergence and improving model performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0030] Figure 1 The figure is a flow chart of a communication optimization federated learning method based on a self-guided evolutionary strategy according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0032] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0033] Example 1

[0034] like Figure 1 As shown, this embodiment provides a communication optimization federated learning method based on a self-guided evolutionary strategy, including:

[0035] The present invention involves two different parties: users and servers, which are described as follows:

[0036] Users: There are multiple users, whose functions are: providing local private datasets that have been trained to implement local model training; using a shared random number seed and the global gradient vectors of the previous k rounds to generate multiple pseudo-random gradient evaluation vectors, and evaluating the fitness of the local gradient vectors and these evaluation vectors, and submitting the fitness to the server; after receiving the global fitness from the server, calculating the global gradient and updating the model parameters. Users may experience disconnection.

[0037] Server: There is a server whose functions are to: initialize the global model parameters and random number seed and send them to each user; after receiving the local fitness of each user, aggregate it into a global fitness. At the same time, use the shared random number seed and the global gradient vector of the previous k rounds to generate multiple pseudo-random gradient evaluation vectors, and use the global fitness calculation to generate the global gradient, update the global model parameters; and send the global fitness to each user.

[0038] Initialization step: The server initializes the global model parameters and random number seed and sends them to each user. After initialization, the global model is continuously updated by looping the following steps, as described below:

[0039] Step (1): The user uses local data training to obtain a local gradient vector, and then uses a random number seed and the global gradient vector of the previous k rounds to divide the full real number space into the principal component space and its orthogonal complement space; pseudo-random evaluation vectors are extracted in a certain proportion from the principal space and the orthogonal complement space, and the fitness value of each evaluation vector with respect to the local gradient vector is calculated. The user sends the local fitness value to the server;

[0040] Step (2): After receiving the local fitness value from each user, the server aggregates it into a global fitness value and sends the global fitness value to each user;

[0041] Step (3): The user and the server calculate the global gradient vector of this round based on the global fitness value and the pseudo-random evaluation vector, and update the global model.

[0042] In this embodiment, after the user completes this round of training locally, a self-guided pseudo-random gradient evaluation vector group is generated using a shared random number seed and the global gradients of the previous k rounds, and the local gradient vectors obtained in this round of training are converted into fitness values ​​for the evaluation vector group. This greatly reduces the communication overhead between the user and the server, and maintains more information than the unguided gradient evaluation vector group, which can ensure faster convergence of the global model training.

[0043] As a preferred implementation, the initialization step specifically includes:

[0044] The server selects the appropriate machine learning / deep learning model based on specific needs and initializes the global model parameters θ 0 , select random number seed s, and set {θ 0 ,s} is sent to each user.

[0045] As a preferred embodiment, step (1) specifically includes:

[0046] In round t, user u uses local training data to train a local model Get the local gradient vector

[0047] If t≤k, user u uses random number seeds to generate a normal distribution The pseudo-random gradient evaluation vector group If t>k, the user performs PCA decomposition on the first k global gradient vectors and takes the first d principal components to form the main space Take the remaining principal components to form an orthogonal complementary space Where d can be determined by taking the logarithm log2() of the dimension of the gradient vector; and sampling in these two subspaces according to the proportion α:

[0048] where i = 1, ..., m;

[0049] Among them, the eigenvectors (principal components) corresponding to the first d eigenvalues ​​are taken according to the size of the eigenvalue.

[0050] In practical applications, in order to balance self-guided sampling and extended sampling, the probability α is usually set to 0.5 or 0.3; secondly, in practical applications, the gradient vector usually has a dimension of more than hundreds of thousands, and storing a matrix of hundreds of thousands by hundreds of thousands dimensions is unrealistic for ordinary users. Therefore, the principal components from d+1 to 2d can be taken to form a pseudo-orthogonal complementary space.

[0051] User u calculates the fitness of each pseudo-random gradient vector and sends it to the server:

[0052]

[0053] in, is the pseudo-random gradient evaluation vector, represents the local gradient vector obtained by user u in the tth round of training, represents the local fitness value of the i-th pseudo-random gradient vector, and m is the number of pseudo-random gradient vectors.

[0054] As a preferred embodiment, step (2) specifically includes:

[0055] The server receives the fitness value vector of each user Then, calculate the global fitness: where w u is the weight of user u;

[0056] The server will set the global fitness value Each user.

[0057] As a preferred embodiment, step (3) specifically includes:

[0058] The server uses the random number seed to generate the same pseudo-random evaluation vector set as the user according to the method described in step (1)

[0059] The server and the user independently calculate the global gradient of this round and update the model parameters:

[0060]

[0061] θ t+1 =θ t -η*g t ;

[0062] The user saves the global gradient g t For future training rounds; where g t is the global gradient obtained by aggregation in the tth round, η is the learning rate, θ t+1 is the t+1th round global model, m is the number of pseudo-random gradient evaluation vectors, is the pseudorandom gradient evaluation vector.

[0063] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A communication optimization federated learning method based on a self-guided evolutionary strategy, characterized in that: The following steps are involved: Step 1: The server initializes the global model parameters and random number seed and sends them to each user; Step 2: The user trains with local data to obtain a local gradient vector; a subspace is divided from the real number space based on the global gradient vector of a preset number of rounds and the random number seed, and a number of pseudo-random evaluation vectors are extracted from the subspace; Calculating the local fitness value of each pseudo-random vector with respect to the local gradient vector, and feeding the result back to the server; Step 3: The server calculates the global fitness value based on the local fitness value fed back by each user and sends it to each user; Step 4: The user and the server calculate the global gradient vector of this round based on the global fitness value and the pseudo-random evaluation vector, and update the global model; Repeat steps 2 to 4 to continuously update the global model.

2. The communication optimization federated learning method based on self-guided evolutionary strategy according to claim 1 is characterized in that: The process of obtaining the pseudo-random evaluation vector in step 2 includes: The user trains the local model using local data to obtain a local gradient vector. If the number of training rounds for the local model is less than or equal to a preset number of rounds, a pseudo-random gradient evaluation vector is generated based on the random number seed. If the number of training rounds for the local model is greater than the preset number of rounds, a subspace is divided from the real number space based on the global gradient vector and the random number seed, and several pseudo-random evaluation vectors are extracted from the subspace.

3. The communication optimization federated learning method based on self-guided evolutionary strategy according to claim 2 is characterized in that: In the step 2, the pseudo-random gradient evaluation vector generated based on the random number seed obeys a normal distribution.

4. The communication optimization federated learning method based on self-guided evolutionary strategy according to claim 2 is characterized in that: In step 2, a subspace is divided from the real number space based on the global gradient vector and the random number seed, and a process of extracting a plurality of pseudo-random evaluation vectors from the subspace includes: The matrix composed of the global gradient vectors of the first few rounds of a preset number of rounds is subjected to matrix decomposition and divided into two parts: the main space and the orthogonal complement space. Sampling is performed in the main space and the orthogonal complement space according to a preset sampling ratio to obtain a number of pseudo-random evaluation vectors.

5. The communication optimization federated learning method based on self-guided evolutionary strategy according to claim 1 is characterized in that: In step 2, the formula for calculating the local fitness value of each pseudo-random vector to the local gradient vector is as follows: in, is the pseudo-random gradient evaluation vector, represents the local gradient vector obtained by user u in the tth round of training, represents the local fitness value of the i-th pseudo-random gradient vector, and m is the number of pseudo-random gradient vectors.

6. The communication optimization federated learning method based on self-guided evolutionary strategy according to claim 5 is characterized in that: In step 3, the formula for calculating the global fitness value is as follows: Among them, w u is the weight of user u, U is the total number of users, f i t is the global fitness value.

7. The communication optimization federated learning method based on self-guided evolutionary strategy according to claim 6 is characterized in that: In step 4, the server generates a pseudo-random evaluation vector set using the method of step 2. The server and the user respectively calculate the global gradient of this round and update the model parameters. The user saves the global gradient for the next round of training.

8. The communication optimization federated learning method based on self-guided evolutionary strategy according to claim 7 is characterized in that: The formula for calculating the global gradient and model parameters of this round is as follows: i t+1 =θ t -h*g t Among them, g t is the global gradient obtained by aggregation in the tth round, η is the learning rate, θ t+1 is the t+1th round global model, m is the number of pseudo-random gradient evaluation vectors, is the pseudorandom gradient evaluation vector.

Images