Semi-automated data classification with feature weighted self organizing map

This paper presents a Feature Weighted Self-Organizing Map (FWSOM) that analyses the topology information of a converged standard Self organizing Map (SOM) to automatically guide the selection of important inputs during training for improved classification of data with irrelevant inputs. We demonstrate an improved classification accuracy with the proposed method by comparison with the standard SOM and other relevant existing classifiers on synthetic and real-world datasets. In addition, the FWSOM method was able to successfully identify the relevant features which in turn were able to improve the classification performance of the other classification methods.