An adaptive strategy for the classification of g-protein coupled receptors

S. Mohamed, D. Rubin, T. Marwala
2007 SAIEE Africa Research Journal  
One of the major problems in computational biology is the inability of existing classification models to incorporate expanding and new domain knowledge. The prohlem of static classification models is addressed in this paper by the introduction of incrcmelllal learning for problems in hioinformatics. Many machine learning 100is have been applied to Ihis problem using static machine learning structun:s such as neural networks or support vector machines that are unable to accoillmodate new
more » ... ion into their existing models. We utilize the fuzzy ARTMAP as an alternate machine learning system that has the ahility of incrementally learning new data as it becomes available. The fuzzy ARTMAP is found to be comparable to many or the widespread machine learning systems. The use of an evolutionary strategy in thc selection and combination of individual classifiers into an cnsemble system, coupled with the incremental learning ability of the fu7.l.y ARTMAP is proven to be suitable as a pattern classifier. The algorithm presentcd is tested using data from the G-Coupled Protein Receptors Database and shows good accuracy of 83%. The system prescnted is also generally applicable, and can he used in problems in genomics and proteomics.
doi:10.23919/saiee.2007.9488130 fatcat:kcn7lxjblbhwbjga4mi7mmmuxy