On the Development of an Empirical Model for Carcinomic Treatment using Chaotic Hyper-Heuristic Algorithm

A new dimension in the field of computational intelligence was introduced in the late nineties to comprehend a vivid combination of several multi-disciplinary areas. The coalescence biology along with data mining and statistical learning have given birth to Bioinformatics that provides various paradigms for studying the behaviour of unknown patterns at the micro level. In the present work, a recently developed human inspired optimization algorithm called search and rescue (SAR) optimization is employed with an improved version of parameters using Chaos theory. CSARO (Chaotic search and rescue optimization algorithm) unlike other existing algorithms has proven to be a better choice for optimising the gene selection mechanism as well as the control parameters of the learning model. This hyper heuristic algorithm obtained by the inclusion of chaos in SAR mainly aims at enhancement of its global search mobility and prevents from getting trapped in the local optimum. A comparative study with other existing techniques on seven benchmark datasets is performed. The performance of the algorithm is tested using evaluation metrics.