Spark implementation of the enhanced Scatter Search metaheuristic: Methodology and assessment

摘要: Abstract Optimization problems arise nowadays in all disciplines, not only the scientific area but also field of engineering or economics, and many others. Currently, challenging optimization require solution methods that consume a significant amount computational resources. The application High-Performance Computing techniques is common approach to obtain efficient implementations traditional parallel computing systems. However, more recent approaches are exploring distributed programming frameworks developed years achieve computations on clusters cloud In this paper we present implementation enhanced Scatter Search metaheuristic using Spark. program was obtained as particularization general software framework support different realisations metaheuristic. aim provide helpful guidance readers interested applying, developing their own, metaheuristics solve Cloud. With twofold objective demonstrating potential parallelization with Spark studying factors influence performance solution, proposal has been thoroughly evaluated two platforms, cluster platform, representative set parameter estimation Computational Systems Biology.