DEVELOPMENT A MULTIFACTORIAL PORTRAIT MODEL OF SOFTWARE COMPLEXES’ SUPPORTING SUBJECTS, USING ARTIFICIAL NEURAL NETWORKS

Impact factors, that are shaping the individualistic perception of the supported objects by the relevant subjects, who interact with them, directly or indirectly, are considered in this research. A form of impact factors’ (performing impact on the supported software complexes) representation has been studied and proposed. Aforementioned form includes: a set of input characteristics of the researched supported object; a set of impact factors in the form of a transformation matrix function; and a set of output characteristics of the resulting perception of the same researched supported object (however in the individualistic perception of each individual subject of interaction with it). The possibility of encapsulation of artificial neural networks inside the aforementioned proposed form (of the supported software complexes’ impact factors representation) was investigated. And the use of multilayer perceptron was proposed and substantiated for the implementation of the appropriate encapsulations. An appropriate multifactorial portrait model of software complexes’ supporting subjects, using artificial neural networks, particularly a multilayer perceptron, has been developed and presented. Also, the applied practical problem of determining the deficient impact factors for each of the software complex’ support team members has been solved. Key words: model, artificial neural networks, multilayer perceptron, impact factors, multifactor portrait, software complexes, support, automation.