Research of approaches and methods of applying artificial intelligence and machine learning in socio-economic processes
Darya Yu. Sakhanevich · Вестник Омского университета. Серия «Экономика» · 2020
One of the problems hindering the development of the socio-economic sphere in the innovative direction is the lack of structuring of approaches and methods used in machine learning as part of the introduction of artificial intelligence (AI) in socio-economic processes. The same problem hinders the growth of the pace of innovative development and, as a result, the improvement of the scientific and technical level of the country. The article classifies and systematizes aspects of machine learning, focuses on the need to accelerate the construction and implementation of algorithms as the basis of AI for increasing the efficiency of managing socio-economic processes. To achieve this goal, the following results are presented: analysis of the concepts of machine learning and AI, study of analytical materials regarding approaches and methods to the introduction of artificial intelligence and prospects for its application in socio-economic processes. There were systematized approaches to machine learning introduction to artificial intelligence depending on the historical period, the implementation of AI, and another, and methods according to the method of machine learning, predictive model data for creating AI algorithms (e.g., probabilistic), and the idea or the nature of the research that uses this technology (assessment and collection of statistical indicators, analysis). The study of the material related to machine learning and AI construction allowed us to draw the following conclusions. The theoretical foundation in the form of mathematical and statistical methods as the basis for building algorithms for creating AI in the framework of machine learning is a necessary part of the process of teaching computers human qualities. However, information about machine learning methods and approaches is mostly scattered, and it is necessary to form a unified methodological base in order to simplify the stage of searching for the right method of creating AI to solve any social, economic or other problem. The presence of such a database will create opportunities to replace one machine learning method for creating AI with another in different fields of activity and socio-economic processes.