Neural network model for developing innovation and investment policy of real economy organisations in conditions of modern digital transformation~i

摘要

The purpose of the study was to suggest and prove the hypothesis that a neural network can make it possible to evolve a neural network model for developing an innovation and investment policy of real economy organizations. The model suggested includes sustainable use of capabilities of three types of artificial intelligence systems based on the Deductor platform, i.e., the Kohonen map, Big Data quantization neural network and AI decision tree system. The research has identified main factors determining a success and dynamic development of innovations. Theoretical foundations of artificial intelligence systems applied in the real sector of the economy and other spheres have been investigated. Theoretical bases for the development of innovation and investment policy have been considered. There has been formed a neural network family that allowed evolving a model for the development of innovative investment policy of organizations. The sales of innovative products have been forecasted. Monographic, design-calculated and artificial intelligence methods have been mainly applied in the research presented. The following results have been obtained. Main factors affecting the production of innovative products in the conditions of market uncertainty have been found. The theoretical foundations of artificial intelligence systems used in various fields of activity have been investigated. A hypothesis has been suggested and proved that a neural network can make it possible to evolve a neural network model for the development of an innovation and investment policy of real economy companies. The model proposed includes sustainable use of capabilities of the neural network family of three types of artificial intelligence systems based on the Deductor platform, i.e., the Kohonen map, BigData quantization neural network and the AI decision tree system. The main conclusions that reflect the results of the study have been formulated and are of great importance for the study of the digital economy problems.