On the Modelling an Artificial Cognitive System Based on the Human-Brain Architecture
The approach to modeling a cognitive system based on the human-brain architecture, called the Natural-Constructive Approach is presented. The key point of this approach is the following: an artificial cognitive system, being a complex multi-level combinat
- PDF / 306,861 Bytes
- 15 Pages / 439.37 x 666.142 pts Page_size
- 27 Downloads / 195 Views
Abstract The approach to modeling a cognitive system based on the human-brain architecture, called the Natural-Constructive Approach is presented. The key point of this approach is the following: an artificial cognitive system, being a complex multi-level combination of various-type neural processors, should be divided into two subsystems, by analogy with two cerebral hemispheres in a human brain. It is shown that one of them should necessarily contain a random element (noise) for generation of information (creativity); it is responsible for learning. The other one, being free of noise, is responsible for memorization and processing the well-known information. Emotions could be interpreted as the noise-amplitude variation and incorporated into the system by coupling the noise amplitude with the additional variable representing the aggregated value of neurotransmitter composition, which reflects the influence of subcortical brain structures. It is shown that the activity of both subsystems should be controlled by the noise-amplitude derivative.
Keywords Generation of information Neuroprocessor Noise amplitude Hemisphere Stimulant Inhibitor
Emotions Learning
1 Introduction The scientific area of Artificial Intelligence (AI) covers various approaches to modeling the cognitive process: the Robotics [1, 2], Active Agent systems [3, 4], neuromorphic (neuron-based) models [5, 6], Brain Re-Engineering (BRE) [7, 8], etc. Let us stress that any neuromorphic model (though this approach seems to be the closest to the goal) inevitably meets with the “explanatory gap” between the brain and the mind [9]. This implies that we do have a lot of information on the O. Chernavskaya (&) D. Chernavskii P.N.Lebedev Physical Institute, Russian Academy of Sciences, Moscow, Russia e-mail: [email protected] Y. Rozhylo NGO Ukrainian Center for Social Data, Kyev, Ukraine © Springer International Publishing Switzerland 2017 K.S. Hale and K.M. Stanney (eds.), Advances in Neuroergonomics and Cognitive Engineering, Advances in Intelligent Systems and Computing 488, DOI 10.1007/978-3-319-41691-5_10
107
108
O. Chernavskaya et al.
single neuron structure and functioning [10, 11], as well as on manifestation of psychological reactions [12–14], but there is lack of idea on how the first could provide the second. Another problem is connected with the Nature challenges: any human-like cognitive model should be able to answer the questions: why there are just two cerebral hemispheres? Why any human person is individual? Why there are men and women? Up to our knowledge, these problems are not in the focus of modern researches, but this is strange. Perhaps, this could be explained by the complexity of the problems of such sort. Indeed, it is very difficult to imagine and create an artificial system which being manufactured according to standard procedure, would be strictly individual. In the papers [15–17], there was proposed and elaborated so called “Natural-Constructive Approach” (NCA) to modeling the cognitive system. This approac
Data Loading...
