A Parallel Bioinspired Framework for Numerical Calculations Using Enzymatic P System With an Enzymatic Environment

Enzymatic numerical P systems are inspired by the biological structure of cells and the “processing of information” regulated by enzymes on chemical objects, where natural numbers are basic entities to work with. Enzymatic numerical P systems can perform arithmetic operations but not complex numerical calculations, such as obtaining suitable values or finding the maximum or minimum element from the data sets. In this paper, a variant of enzymatic numerical P systems named an enzymatic numerical P system with an enzymatic environment is proposed, which can make numerical operations more flexible with regulations of external enzymes. We design several parallel computational frameworks for complex numerical calculations by the enzymatic numerical P system with an enzymatic environment. A parallel framework is achieved for performing support vector machine (SVM) calculations and sequential minimal optimization by the enzymatic numerical P system with an enzymatic environment. The experimental results obtained using a GPU prove that our method is 5.34 times more efficient than the traditional serial SVM calculation, without any loss in accuracy.