METAHEURISTIC METHODS BASED ON THE BEHAVIOUR OF SOCIAL SPIDERS FOR INTERNAL AUDIT TASKS

Abstract

Today, the actual task is to increase the efficiency of the internal audit of the network production activities and trade systems in terms of IT application. To ensure the efficiency of such systems operation, when decision-making on the choice of assortment, volume and purchase (sale) price, optimization methods are developed. Optimization methods, which find the exact solution, have high computational complexity. Optimization methods, that find an approximate solution using directed search, have a high probability of hitting a local extremum. Random search methods do not guarantee convergence. In this connection, there is an optimization method insufficient efficiency problem, which needs to be solved. Today, these problems are solved by using metaheuristic optimization methods, for example, based on the behavior of social spiders. Classical methods based on the behavior of social spiders do not consider the iteration number in the solution generation operator, which reduces the accuracy of the solution search. The methods proposed in work allow to eliminate the mentioned shortcomings. To find the minimum continuous functions, a population metaheuristic method based on social spiders optimization was developed. Due to the dynamic parameters use, the proposed method performs a global search at the initial iterations and a local search at the final iterations, allowing to an increase in the speed and accuracy of the search. To find the minimum continuous functions, a population metaheuristic method based on the social spider algorithm was developed. Due to the dynamic parameters use, the proposed method performs a global search at the initial iterations and a local search at the final iterations, allowing to increase in the speed and accuracy of the search. The proposed optimization methods based on population metaheuristics can be used to identify parameters of artificial neural networks of information models for audit data transformation. Prospects for further research consist in testing the proposed methods on a wider set of test data. The proposed methods numerical study was carried out on the Ackley function example.