Local repairing method of machine learning model

Abstract

The present invention provides a local repairing method of a machine learning model. The method comprises: collecting of analysis feedback data: collecting user feedback data, and extracting mispredicted data samples; performing spatial alternation: converting an original data space to a new data space through scale learning, reducing a distance between the mispredicted data samples as much as possible in the new data space, and increasing a distance between the mispredicted data samples and correctly predicted data samples as much as possible; learning incorrect data samples to establish a patch model in the new data space, and defining the application range of the patch model; and in the new data space, learning incorrect data samples to establish the patch model, and defining the application range of the patch model. The local repairing method of the machine learning model improves the performance of the machine learning model.

Description

technical field [0001] The invention relates to a local repair method of a machine learning model, which belongs to the field of Internet search. Background technique [0002] With the rapid development of the Internet, search engines have become an important tool for people to use Internet information resources. With Google, Yahoo! With the rise and development of search engines such as , Bing, and Baidu, the relevance of query results has attracted more and more attention. The ranking of query results has also become the main index for evaluating search engines. [0003] With the rapid development and wide application of information technology, the Internet has developed vigorously and become the world's largest information resource, which has occupied an important position in people's lives. The Internet has also become an important platform for people to share and interact with each other. It is like finding a needle in a haystack for users to find the information th...

AI Technical Summary

Problems solved by technology

The learning model established by re-learning is mainly determined by the original training data set, so the room for performance improvement is very limited

[0008] 2. Every time a small amount of user feedback is obtained, it is necessary to relearn and build a new machine learning model, which will inevitably greatly reduce the effect of search

Images