Solving optimization problems by using networks of evolutionary processors with quantitative filtering

Searching for new efficient algorithms to solve complex optimization problems in big data scenarios is a priority, especially when the search space increases exponentially with the problem size, making impossible to find a solution through a mere blind search. Networks of Evolutionary Processors (NEP) is a formal framework formed of highly parallel and distributed computing models inspired and abstracted from biological evolution that is able to solve hard problems in an efficient way. However, NEP is not expressive enough to model quantitative aspects present in many problems. In this paper we propose NEPO, a new model based on the NEP evolutionary processors. NEPO deals with a class of data that is able to solve hard optimization problems and defines a novel selection process based on a quantitative filtering strategy. We present a linear time solution to a well known NP-complete optimization problem (the 0/1 Knapsack problem) in order to demonstrate NEPO advantages. This result suggests that NEPO's quantitative filtering is more suitable to tackle practical solutions to optimization problems in order to deploy them on highly scalable distributed computational platforms.