Neural networks in training and analysis of complex data in the educational process

Abstract

Changes in economic processes and indicators have a significant impact on each individual and humanity as a whole. If we analyze the activities of any person, we will notice that economic issues will occupy leading places in his life. Therefore, it is extremely important to analyze economic processes quickly and at the same time spend a minimum of intellectual resources and reduce the level of error. With the development of information science and the observation of the structure of our brain, neural networks have emerged that can significantly speed up data analysis and reduce error. And also to find in the processes or phenomena that are being analyzed, nonlinear dependencies, patterns and cause-and-effect relationships that may not be taken into account when narrowly considering the data. Neural networks show themselves to be the most effective when working with a large amount of data that a person is physically unable to perceive. For example, today everyone uses search services, electronic payment systems and smartphones with geolocation, so artificial neural networks are widely used by corporations to find which product will interest each particular buyer, or for personalized advertising based on data obtained about the user (if this data is available). However, it can be said that neural networks are not a perfect tool and have a number of disadvantages, since there are many architectures designed to solve individual tasks and require a long learning process and a significant amount of data. But with the right choice of network type and its further configuration and training, you can achieve exceptional results in analyzing a large amount of data. That is why the study of the topic of the use of neural networks in economics is an urgent and necessary scientific direction that requires more in-depth research.