Federal learning algorithm component priority scheduling implementation method and device and storage medium

Abstract

The embodiment of the invention discloses a federated learning algorithm component priority scheduling implementation method and device and a storage medium, and the method comprises the steps: obtaining an execution process of an algorithm component, analyzing the execution process of the algorithm component into a directed acyclic graph, and carrying out the inversion of the directed acyclic graph, generating a reverse directed acyclic graph; performing topological sorting on the reverse directed acyclic graph to generate an execution sequence; the execution sequences are traversed, the execution weight of each node is calculated, the execution weight is the sum of the node self weight of the node and the node in-degree weight of the node, the execution sequences are arranged in a descending order based on the size of the execution weights, and the priority scheduling sequence of the algorithm components corresponding to the nodes is obtained. According to the method, the execution weight is calculated according to the generated execution sequence, and the execution sequence of the algorithm components is defined according to the execution weight, so that automatic scheduling among the algorithm components according to the priority is realized.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a method, device and storage medium for implementing priority scheduling of components of a federated learning algorithm. Background technique [0002] Today, when the data is increasing day by day and the data is becoming more and more closely connected, due to user privacy, laws and regulations and other issues, many data cannot be communicated with each other, resulting in many "data silos". The concept of federated learning (Federated Learning) was proposed by Google in 2017 to solve the joint modeling problem between cross-devices. This scheme provides a feasible solution to the above problems. During the modeling process, modelers usually need to go through processes such as data reading and writing, data preprocessing, statistical analysis, feature engineering, machine learning, prediction, and evaluation. On the federated modeling platform, these operations are...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the embodiments of the present invention is to provide a method, device, and storage medium for implementing priority scheduling of federated learning algorithm components to solve the problem that the existing technology cannot customize the execution order of algorithm components, thus resulting in the inability to implement priority scheduling among algorithm components. Problems with Automated Scheduling

Images