Hybrid model for the efficient formation of training courses personalized recommendations

Abstract

In this study, a hybrid model that combines collaborative filtering and deep neural networks is presented to improve the effectiveness of learning recommendations. This approach allows to optimise the recommendation generation process, providing high flexibility and accuracy even with a limited amount of data. In cases where the available data is not enough to train the neural network efficiently, the algorithm automatically switches to collaborative filtering, which preserves the quality of recommendations and the stability of the model. The results of testing on datasets have shown that the hybrid model improves the accuracy and adaptability of the recommendation system compared to traditional methods. The combination of the two methods allows to use their advantages and minimise their disadvantages, making the model versatile in different contexts and with different amounts of data. This is especially relevant in the field of education, where user data may be limited or incomplete. Implementation of this model makes it possible to create more effective educational platforms that provide personalised learning for a wide audience, taking into account the individual needs and preferences of users. An important feature of the algorithm is its ability to adapt to changes in learning content and customise recommendations according to the needs of each user. This helps to increase user satisfaction and improves the overall quality of educational services. Effective work with limited data increases the accuracy and adaptability of the recommendation system. All of these properties make the proposed algorithm a promising tool for developing personalised educational platforms that better meet individual needs. This contributes to more efficient learning and better assimilation of the material, which is important for the development of the modern information society