Neural network model training method and device, and computer device

Abstract

A neural network model training method and device, and a computer device are disclosed. The neural network model training method includes: performing training by using a source domain sample set to obtain a neural network model; performing iterative processing by using the following steps until an iterative stop condition is satisfied: inputting an uncertain label sample in a target domain sampleset into the current neural network model, and using the output prediction result as a current learning label of the uncertain label sample; inputting a source domain sample set and the current targetdomain sample set in to the current neural network model, constructing a neural network loss function on the basis of the output prediction result, wherein independent variables of the neural networkloss function include at least: a difference between a mapping feature of the source domain sample set and a mapping feature of the current target domain sample set; using the current neural networkloss function to perform back propagation so as to update model parameters of the current neural network model; and determining the current neural network model as the final trained neural network model after iteration is ended.

Description

technical field [0001] The embodiments of this specification relate to the technical field of data processing, and in particular to a training method, device, and method for a neural network model. Background technique [0002] In machine learning, deep learning, data mining and other tasks, a large number of data samples can be used for training, and various forms of data models can be obtained to solve practical problems. At the same time, in these tasks, the ideal situation is that all data samples are With labels, all data samples obey the same distribution and come from the same feature space, so that the data model can be obtained through supervised learning. When the performance of the trained data model meets the requirements, the data model can be put into use To classify and predict unknown data. [0003] However, in practical applications, it may happen that: on the one hand, due to the limitation of the application scenario, there is a lack of sufficient labeled...

AI Technical Summary

Problems solved by technology

[0003] However, in practical applications, it may happen that: on the one hand, due to the limitation of the application scenario, there is a lack of sufficient labeled training samples, which makes the model training more difficult; on the other hand, for the same application scenario, As time goes by, the data distribution will change with the change of some dynamic factors, resulting in that the characteristics of the unknown data and the data samples used in training the model are not exactly the same, so the existing data model is used to carry out these unknown data. When predicting, the expected effect will not be achieved. At the same time, the previously collected data samples will no longer be applicable, and data samples need to be collected again to retrain the model

Images