This article presents a decision support algorithm for the commercialization of intellectual property that integrates regression analysis and machine learning methods. The algorithm takes a range of factors into account, including patent features, market indicators, technological trends, and economic conditions. A formalized problem statement with an objective function for minimizing the mean square error is proposed, and the algorithm implementation stages are detailed: from data collection and preprocessing to the construction, validation, and dynamic updating of the predictive model. Particular attention is paid to the implementation of a dynamic assessment mechanism for technological trends to improve the model’s adaptability.
Prudnikov et al. (Sun,) studied this question.