Generalization of metric classification algorithms for sequences classification and labelling

TL;DR

This paper proposes a generalized metric classification algorithm tailored for sequential data, specifically extending k-Nearest Neighbours, and demonstrates its effectiveness against CRF in chunking tasks.

Contribution

It introduces a novel generalization method for metric classification algorithms applied to sequence labeling, including a new classification and labeling algorithm.

Findings

The proposed algorithm outperforms existing methods in certain sequence classification tasks.

Experimental results show competitive accuracy with CRF in chunking on CoNLL2000.

The method offers advantages in simplicity and effectiveness for sequential data classification.

Abstract

The article deals with the issue of modification of metric classification algorithms. In particular, it studies the algorithm k-Nearest Neighbours for its application to sequential data. A method of generalization of metric classification algorithms is proposed. As a part of it, there has been developed an algorithm for solving the problem of classification and labelling of sequential data. The advantages of the developed algorithm of classification in comparison with the existing one are also discussed in the article. There is a comparison of the effectiveness of the proposed algorithm with the algorithm of CRF in the task of chunking in the open data set CoNLL2000.