A multi-instance learning algorithm based on a stacked ensemble of lazy learners
This incremental work addresses classification problems where instance-level labels are unavailable or imprecise, such as in drug discovery or image analysis.
The paper tackles multi-instance learning by developing a stacked ensemble algorithm using lazy learners to classify bags of instances, achieving effectiveness on the Musk1 benchmark dataset.
This document describes a novel learning algorithm that classifies "bags" of instances rather than individual instances. A bag is labeled positive if it contains at least one positive instance (which may or may not be specifically identified), and negative otherwise. This class of problems is known as multi-instance learning problems, and is useful in situations where the class label at an instance level may be unavailable or imprecise or difficult to obtain, or in situations where the problem is naturally posed as one of classifying instance groups. The algorithm described here is an ensemble-based method, wherein the members of the ensemble are lazy learning classifiers learnt using the Citation Nearest Neighbour method. Diversity among the ensemble members is achieved by optimizing their parameters using a multi-objective optimization method, with the objectives being to maximize Class 1 accuracy and minimize false positive rate. The method has been found to be effective on the Musk1 benchmark dataset.