Affiliation:
1. Siberian Federal University
Abstract
Abstract. The authors construct a groupoid whose elements are associated with multilayer feedforward neural networks. This groupoid is called the complete groupoid of the composition of neural networks. Multilayer feedforward neural networks (hereinafter referred to as neural networks) are modelled by defining a special type of tuple. Its components define layers of neurons and structural mappings that specify weights of synaptic connections, activation functions and threshold values. Using the artificial neuron model (that of McCulloch-Pitts) for each such tuple it is possible to define a mapping that models the operation of a neural network as a computational circuit. This approach differs from defining a neural network using abstract automata and related constructions. Modeling neural networks using the proposed method makes it possible to describe the architecture of the network (that is, the network graph, the synaptic weights, etc.). The operation in the full neural network composition groupoid models the composition of two neural networks. A network, obtained as the product of a pair of neural networks, operates on input signals by sequentially applying original networks and contains information about their structure. It is proved that the constructed groupoid is a free.
Publisher
National Research Mordovia State University MRSU
Reference8 articles.
1. V.A. Golovko, V.V. Krasnoproshin, Neural network technologies for data processing, Belarus. State University Publ., Minsk, 2017 (In Russ.), 263 p.
2. A.N. Gorban, “Generalized approximation theorem and computational capabilities of neural networks” , Siberian Journal of Computational Mathematics, 1:1 (1998), 11–24 (In Russ.).
3. F. Rosenblatt, Principles of neurodynamics: the perceptron and the theory of brain mechanisms, World, Moscow, 1965 (In Russ.), 478 pp. p.
4. A.V. Sozykin, “Review of methods for training deep neural networks” , Bulletin of the South Ural State University. Series “Computational Mathematics and Informatics”, 6:3 (2017), 28–59 (In Russ.). DOI: https://doi.org/10.14529/cmse170303
5. A.V. Litavrin, “Endomorphisms of finite commutative groupoids associated with multilayer feed-forward neural networks” , Tr. IMM Ural Branch RAS, 27:1 (2021), 130–145 (In Russ.). DOI: https://doi.org/10.21538/0134-4889-2021-27-1-130-145