SOLVING CHALLENGES IN BUSINESS STRUCTURES WITH THE HELP OF GENERATIVE MODELS

For the business models addressed in this study, we propose the implementation of distribution free learning framework concepts and paradigms. The development of a machine learning process for a predictor identified with a high degree of precision is done with the help of a discriminatory paradigm. A generative-type approach is developed, using the hypothesis that the underlying distribution used for the sampled and interpreted data has a parametric structure exploiting the so-called parametric density estimation. This choice has the advantage of avoiding learning processes for the distributions underlying the business models, resulting in rigorous predictions. For the economic models, we consider that the VANIK principle has a relevant degree of efficiency, using a well-defined amount of information. The originality and solutions proposed in this work come from the idea that in order to manage economic organizations, we must turn to innovative technological concepts and paradigms, such as machine and deep Learning as part of Artificial Intelligence. Therefore, economic activities will have both a controlled degree of uncertainty and a high degree of operational-strategic performance.