Probability domain generalization learning method based on meta-learning

Abstract

The invention discloses a probability domain generalization learning method based on meta-learning, belongs to the field of meta-learning, and aims to combine a meta-learning thought into domain generalization for the first time and solve the problem that parameters are linearly increased along with increase of the number of source domains in domain generalization by utilizing a meta-learning framework. The variational information bottleneck idea is combined into meta-learning and domain generalization for the first time, so that the generalization ability of the patent can be further improved; according to the method, the problem that parameters linearly increase along with the number of source domains can be solved through meta-learning; according to the technical scheme, the probabilitydomain generalization learning method based on meta-learning is formed by combining the variational thought with the information bottleneck and fusing the variational thought and the information bottleneck into a unified probability framework, and the brand-new and effective probability domain generalization learning method based on meta-learning can be formed by combining the variational thoughtwith the information bottleneck.

Description

technical field [0001] The invention relates to the field of meta-learning, and more specifically, relates to a meta-learning-based generalization learning method in probability domain. Background technique [0002] Traditional machine learning assumes that training data and test data obey the same data distribution, which is difficult to meet in practical applications. There are several classical approaches to this problem, including 1) transfer learning: the goal of transfer learning is to use knowledge learned from one environment to help learning tasks in a new environment; 2) domain adaptation: the focus of domain adaptive learning It lies in how to overcome the difference between the distribution of the source domain and the distribution of the target domain, and realize the learning task on the target domain; 3) domain generalization: when the target domain is unknowable, the distribution or model has good characteristics for unknown situations. The difficulty of the...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the problems existing in the prior art, the purpose of the present invention is to provide a probabilistic domain generalization learning method based on meta-learning, which can realize the first combination of meta-learning ideas into domain generalization, and use meta-learning framework to solve domain The problem that the parameters increase linearly with the increase of the number of source domains; the first time the idea of ​​variational information bottleneck is combined with meta-learning and domain generalization, which can further increase the generalization ability of this patent; this scheme can solve the problem of parameter randomization through meta-learning Aiming at the problem of linear increase in the number of source domains, and through the meta-learning framework, domain-invariant feature representations can be obtained more accurately. In order to further increase the domain generalization performance of this scheme, this scheme combines variational thinking and information bottlenecks, and integrates its Integrate into a unified probability framework to form a new and extremely effective meta-learning-based probability domain generalization learning method

Images