Assessment of influence productivity in cognitive models
- PDF / 1,271,019 Bytes
- 27 Pages / 439.37 x 666.142 pts Page_size
- 97 Downloads / 194 Views
Assessment of influence productivity in cognitive models Alexander Tselykh1 · Vladislav Vasilev1 · Larisa Tselykh2
© Springer Nature B.V. 2020
Abstract This article proposes a new influence productivity assessment methodology that is a cognitive intelligence system for the scenario planning of control impacts (generation and choice) for systems that are represented by directed weighted signed graphs based on the algorithm of effective controls. The algorithm implements a control model that expresses the direction of development (growth) of the system. The algorithm is based on the spectral properties of the adjacency matrix of a graph representing the model of a socioeconomic system and does not impose any constraints on the directions of the edges or the sign and weight range on the edges. Scenarios are assessed based on their compliance with tactical and strategic goals according to the codirectionality degree of the response vector with respect to the base vector of the model. The base vector is the effective control vector without constraints on the controls under the conditions of adequate model operation. The new methodology has three distinctive features: (1) the scenario approach is implemented with respect to a set of controls, (2) this approach is applicable for models with heterogeneous factors and does not require preliminary aggregation of the primary model elements of the system; and (3) this approach has a clear formalization metric for the selecting and generating of a set of control impacts. The process does not require the decision maker to have special mathematical training. Keywords Influence productivity · Directed weighted signed graphs · Cognitive model · Optimization methods · Complex systems
1 Introduction This paper addresses the problem of robotizing the decision-making process for choosing control impact scenarios for models of socioeconomic systems represented by directed weighted signed graphs. The most widespread graph forms of such systems are networks. Among the network models, we consider a nongrowing causal model with a fixed number * Larisa Tselykh [email protected] 1
Department of Information and Analytical Security Systems, Institute of Computer Technologies and Information Safety, Southern Federal University, Nekrasovskii, 44, Taganrog, Russia 347900
2
Department of Economics and Business, Chekhov Taganrog Institute, Rostov State University of Economics, Initsiativnaya, 48, Taganrog, Russia 347900
13
Vol.:(0123456789)
A. Tselykh et al.
of network nodes. The basic idea underlying the solution to this problem is the person’s ability to generalize and abstract when making decisions. The decision maker has his own frame of reference, which includes (1) an image of the factors and relationships in the system, (2) an image of the system as a whole and the general direction of the development of the system, and (3) their own judgments about the main factors of the system, which, when altered, will change the state of the system. Based on this, a person makes deci
Data Loading...
