A Semi-Autonomous Online Learning Method Based on Random Fern Classifier

Abstract

The invention provides a semi-autonomous online learning method based on a random fern classifier, comprising the following steps: 1) preparing a sample set for initial training classifier; 2) initial training of a random fern classifier; 3) obtaining online learning samples; 4) random fern classifier Online training of classifiers. The present invention proposes an online learning algorithm of a random fern classifier so that the target detection system can gradually improve its performance and finally meet the detection accuracy requirements; in addition, the method of obtaining online learning samples through manual judgment ensures the correctness of online learning sample category labeling; Compared with the traditional classifier training method, the present invention does not need to prepare a large number of positive and negative training samples in advance, reduces the workload of manual labeling, and can verify the classification performance of the random fern classifier.

Description

technical field [0001] The invention belongs to a pattern recognition method, and in particular relates to a method for training a classifier through an online learning algorithm to improve the performance of the classifier. Background technique [0002] Online learning belongs to the research category of incremental learning. In this type of method, the classifier only learns each sample once instead of repeated learning. In this way, a large amount of storage space is not required to store training samples during the operation of the online learning algorithm. Every time the classifier obtains a sample, it conducts online learning. Through online learning, the classifier can still update and improve itself according to the new sample during use, and further improve the classification effect. [0003] Early online learning algorithms include Winnow algorithm, unified linear prediction algorithm, etc. In 2001, scholar Oza combined these algorithms with boosting algorithm and...

AI Technical Summary

Problems solved by technology

However, online Adaboost uses feature selection operators instead of general weak classifiers to synthesize strong classifiers. The number of feature selection operators and the number of weak classifiers corresponding to feature selection operators are fixed, and the corresponding online learning classifier structure is relatively rigid.

When it is found that its classification ability cannot meet the requirements of detection performance, even continuous online learning cannot improve the detection accuracy

Images