Robust Classification with Context-Sensitive Features

TL;DR

This paper introduces strategies for improving classification accuracy when features are influenced by changing contexts, demonstrated across domains like engine diagnosis, speech recognition, and medical prognosis.

Contribution

It provides a formal definition of context-sensitive classification and presents general methods to enhance classifier performance in varying contexts.

Findings

Context-aware strategies improve classification accuracy across domains.

Significant performance gains observed in engine diagnosis, speech recognition, and medical prognosis.

Methods are effective when training and testing contexts differ.

Abstract

This paper addresses the problem of classifying observations when features are context-sensitive, especially when the testing set involves a context that is different from the training set. The paper begins with a precise definition of the problem, then general strategies are presented for enhancing the performance of classification algorithms on this type of problem. These strategies are tested on three domains. The first domain is the diagnosis of gas turbine engines. The problem is to diagnose a faulty engine in one context, such as warm weather, when the fault has previously been seen only in another context, such as cold weather. The second domain is speech recognition. The context is given by the identity of the speaker. The problem is to recognize words spoken by a new speaker, not represented in the training set. The third domain is medical prognosis. The problem is to predict whether a patient with hepatitis will live or die. The context is the age of the patient. For all three domains, exploiting context results in substantially more accurate classification.