Lateral federated gradient boosting tree optimization method based on random greedy algorithm

Abstract

The invention discloses a transverse federated gradient boosting tree optimization method based on a random greedy algorithm, which comprises the following steps that: a coordinator sets related parameters of a gradient boosting tree model, including but not limited to the maximum number T of decision trees, the maximum depth L of the tree, an initial prediction value base and the like, and issues the parameters to each participant pi; and each participant segments the current node data set according to the segmentation feature f and the segmentation value v, and allocates new segmentation data to the child nodes. According to the transverse federated gradient boosting tree optimization method based on the random greedy algorithm, supported transverse federated learning comprises the participant and a coordinator, the participant has local data, and the coordinator has local data. The coordinator does not have any data, the center and the participants performing information aggregation of the participants respectively calculate histograms and send the histograms to the coordinator, and the coordinator gathers all histogram information, searches an optimal break point according to a greedy algorithm, and then shares the optimal break point to each participant to work in cooperation with an internal algorithm.

Description

technical field [0001] The invention relates to the technical field of federated learning, in particular to a horizontal federated gradient boosting tree optimization method based on a random greedy algorithm. Background technique [0002] Federated learning is a machine learning framework that can effectively help multiple institutions to conduct data usage and machine learning modeling while meeting the requirements of user privacy protection, data security, and government regulations, allowing participants to jointly build on the basis of unshared data. Model, which can technically break the data island and realize AI collaboration. Under this framework, the problem of different data owners collaborating without exchanging data is solved by designing a virtual model. The virtual model is the aggregation of data by all parties. The optimal model, each area serves the local target according to the model, and federated learning requires that the modeling result should be inf...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a random greedy algorithm-based horizontal federated gradient boosting tree optimization method to solve the problem that the existing horizontal federated gradient boosting tree algorithm in the above-mentioned background technology requires each participant and coordinating party to frequently transmit histogram information, The network bandwidth requirements of the coordinator are very high, and the training efficiency is easily affected by network stability. Since the transmitted histogram information contains user information, there is a risk of leaking user privacy. When introducing multi-party secure computing, homomorphic encryption, and secret sharing After waiting for the privacy protection scheme, the possibility of user privacy leakage can be reduced, but it will increase the local computing burden and reduce the training efficiency

Images