Federal decision tree construction and prediction method, system and device, and storage medium

Abstract

The embodiment of the invention discloses a federal decision tree construction and prediction method and system, equipment and a storage medium. The federated decision tree prediction method comprises the following steps: a second participant determines classification information of a to-be-predicted sample in a specified layer based on a split point of the specified layer in a target federated decision tree stored by the second participant and a second classification feature of the to-be-predicted sample, and sends the determined classification information to a first participant; the first participant determines classification information of the to-be-predicted sample in other layers based on splitting points of other layers except the specified layer in the target federated decision tree and the first classification feature of the to-be-predicted sample, and determining a prediction result of the to-be-predicted sample according to the target federal decision tree, the classification information of the to-be-predicted sample in the other layers and the classification information of the to-be-predicted sample in the specified layer sent by the first participant. According to the technical scheme of the embodiment of the invention, the prediction efficiency can be improved, and the occupation of network bandwidth and computing resources is reduced.

Description

technical field [0001] The present application relates to the field of computer technology, and in particular, to a method, system, device, and storage medium for constructing and predicting a federated decision tree. Background technique [0002] As a new type of machine learning technology, federated learning can effectively help multiple parties perform data sharing and machine learning modeling while meeting the requirements of user privacy protection, data security, and government regulations, opening up connections between multiple parties. data barriers. Among them, multiple participants can build a federated tree model based on federated learning, and perform data prediction based on the federated tree model. However, in the related art, the data prediction efficiency based on the federated tree model is low, and the amount of data transmission between the participants is large, and the network bandwidth and computing resource occupancy are high, which cannot meet t...

AI Technical Summary

Problems solved by technology

However, in related technologies, the data prediction efficiency based on the federated tree model is low, and the amount of data transmission between participants is large, and the network bandwidth and computing resource usage are high, which cannot meet the requirements of high concurrency, high throughput, and low latency.

Images