Learning from String Sequences

TL;DR

This paper demonstrates that using the Universal Similarity Metric (USM) with a K-NN classifier improves pattern recognition accuracy on sequence data across domains like spam filtering and protein localization.

Contribution

It introduces the USM as an effective alternative distance metric for sequence classification, outperforming traditional string-to-word vector methods.

Findings

USM-based K-NN yields higher accuracy than vector approaches

USM can generate reliable probability forecasts

Effective across diverse sequence domains

Abstract

The Universal Similarity Metric (USM) has been demonstrated to give practically useful measures of "similarity" between sequence data. Here we have used the USM as an alternative distance metric in a K-Nearest Neighbours (K-NN) learner to allow effective pattern recognition of variable length sequence data. We compare this USM approach with the commonly used string-to-word vector approach. Our experiments have used two data sets of divergent domains: (1) spam email filtering and (2) protein subcellular localization. Our results with this data reveal that the USM-based K-NN learner (1) gives predictions with higher classification accuracy than those output by techniques that use the string-to-word vector approach, and (2) can be used to generate reliable probability forecasts.