Drug/nondrug classification with consensual Self-Organising Map and Self-Organising Global Ranking algorithms

Abstract

In this paper, a special consensual approach is discussed for separating the druglike compounds from the non-druglike compounds. It involves a group decision to produce a consensus of multiple classification results obtained with a single classification algorithm. The individual results are obtained with either the Self Organising Global Ranking (SOGR) or Self Organising Map (SOM). The main difference between the proposed algorithm and SOM is the neighbourhood concept. The constructed consensual model has a preprocessing unit which consists of transformation of input patterns by random matrices and median filtering to generate independent errors for a single type of classifier, and a postprocessing unit for consensus. The confirmed drugs were classified with a consensual accuracy of 90.63% while nondrugs resulted in 80.44% accuracy. The SOGR results were better than the SOM algorithm results.