CREATION OF LOGOS USING MACHINE LEARNING METHODS

The article explored the features of using neural networks to create logos. The significant influence of globalization processes and digitalization on the activation of marketing strategies development in the offline and online environment has been proven. Prerequisites for the active development of machine learning algorithms, including neural networks with different architectures and activation functions, are given. The expediency of establishing long-term communications between the company and the target audience thanks to the use of logos has been established. The effectiveness of different color schemes use in the development of logos, taking into account the set of the target audience characteristics, was determined. It has been established that the main customers who pay attention to innovative logos are representatives of the Z and Alpha generations. Communication between the company and the target audience is carried out thanks to the stimulation of various senses. Taking into account biological features, consumers perceive the main information thanks to the organs of vision. Accordingly, the creation of a high-quality logo increases the probability of the company's success on the market, but only if a comprehensive marketing strategy is implemented in the offline and digital environment. A correctly selected font affects the perception of the logo by consumers. Choosing the wrong typeface hinders the communication of the brand's philosophy to consumers. Along with this, the use of specific fonts can lead to complications when reading the text on the logo. Individual fonts are negatively perceived by customers and can lead to a decrease in sales volumes. To identify the company, it is advisable to place the logo on banners, business cards, in advertising, in social networks, etc. The logo must be light, elegant and unique to ensure ease of perception of the graphic object as an integral part of the corporate identity of the company. A typical architecture of a neural network with a description of the main layers of input and processing of large volumes of information is presented. There are examples of software that allows companies to generate visual content, including logos for various brands, based on text information.