APPLICATION OF TRANSFORMER-ARCHITECTURE ARTIFICIAL INTELLIGENCE MODELS IN MANAGING THE DEVELOPMENT OF SPEED ENDURANCE IN FINSWIMMERS AT THE STAGE OF ADVANCED SPECIALIZATION
Back to articles of current journal| PDF: |  |
Author(s): | Dudchenko P. P., |
| Number of journal: | 2(75) | Date: | June 2026 |
| Annotation: | The article is devoted to the approbation of a digital analytical framework based on Transformer-architecture artificial intelligence models in training 15–16-yearold finswimmers at the stage of advanced specialization. The focus is on integrating an intelligent coach assistant into the planning and ongoing correction of training means aimed at developing speed endurance. The approbation of the digital analytical framework based on a Transformer-architecture artificial intelligence model was conducted in 2025 at the Tula Regional Comprehensive Sports School of Olympic Reserve. The study involved 30 finswimmers with at least eight years of training experience and qualifications of First Class and Candidate for Master of Sports. Based on their initial 100 m performance, the athletes were divided into a control group (CG, n = 15) and an experimental group (EG, n = 15). In the EG, the coach used digital tools, including an analytical module based on a Transformer model, which processed training logs, split times, heart rate indicators, subjective load assessments, and brief text reports from the athletes. Based on this information, recommendations were generated regarding the microcycle structure, the density of speed-endurance work, and the nature of correction of key training tasks. The effectiveness of the approach was assessed by the 100 m finswimming result, the time of the final 25 m segment, the speed decrement index in the 6×25 m test, lnRMSSD indicators, and the rating of perceived exertion. After the completion of the program, the EG showed more pronounced improvements in sportspecific indicators. The 100 m time decreased by 3.1%, the final 25 m time decreased by 4.9%, and the speed decrement index decreased by 15.7%, while the recovery background proved to be more favorable. The obtained data indicate that incorporating a Transformer model into the analytical support of the coach increases the accuracy of load individualization and contributes to more stable speed realization in the second half of the race. | ||
| Keywords: | finswimming, speed endurance, advanced specialization stage, digital technologies in sporst, Transformer models, intelligent coach assistant, training process management, sports training, load individualization, heart rate variability, rating of perceived exertion / RPE |
||
| For citation: | Dudchenko P. P. Application of Transformer-architecture artificial intelligence models in managing the development of speed endurance in finswimmers at the stage of advanced specialization. Biznes. Obrazovanie. Pravo = Business. Education. Law. 2026;2(75):404—409. DOI: 10.25683/VOLBI.2026.75.1605. |
||