Decision Support System for Personalized Gym Training Planning
Students Name: Palii Bohdan Yuriiovych
Qualification Level: magister
Speciality: Systems and Methods of Decision Making
Institute: Institute of Computer Science and Information Technologies
Mode of Study: full
Academic Year: 2025-2026 н.р.
Language of Defence: ukrainian
Abstract: The active development of mobile fitness applications has opened up wide possibilities for workout personalization. Modern technologies, particularly artificial intelligence and neural networks [1], as well as the availability of wearable devices [5], create ideal conditions for a deep analysis of a person’s physical state. However, common solutions on the market, such as MyFitnessPal [2] or Freeletics [3], are often limited to general recommendations or focus primarily on nutrition tracking. The key problem lies in the lack of deep adaptation of training plans to the user’s unique needs, especially in the context of safety. Existing systems rarely account for medical contraindications or specific limitations, which can lead to ineffective or even dangerous loads, increasing the risk of injury. Thus, an urgent need has arisen for the development of an intelligent system capable of offering a truly individualized and safe plan. Within this master’s qualification thesis, a solution to this problem was proposed. The aim was to create a system that automates program selection based on comprehensive data analysis. To justify the architecture, a system analysis was conducted. The application of the Analytic Hierarchy Process (AHP) method proved that the Decision Support System (DSS) format is the most appropriate for solving the set tasks, receiving the highest priority (0.32) among alternatives [8]. The conceptual model and component interaction logic were designed using the standard UML modeling language [10]. The designed system is based on a hybrid personalization method [9]. This approach combines content-based filtering (for the initial selection of exercises according to goals) with a set of expert rules (to ensure safety and account for medical restrictions). This architecture also lays the foundation for the future integration of deep learning modules [4] for dynamic plan correction based on progress. To support interaction with trainers, features are included, which is a current market trend [7]. For the software implementation, the "Momentum" prototype was created as a single-page web application (SPA) based on the React library. Its functionality was confirmed through control cases: the system demonstrated correct plan generation according to the goal (Case 1) and successfully adapted the plan for a medical restriction, replacing a dangerous exercise with a safe alternative (Case 2). Object of research – the process of planning individual gym workouts using IT solutions. Subject of research – methods and algorithms for analyzing a user’s physical parameters, enabling the automatic formation of training programs based on this data. Aim of the research – the development of a decision support system to personalize the gym training process based on the analysis of the user’s physical parameters and other factors, including integration with medical data [6]. The result of the research is a working prototype of the "Momentum" DSS, which solves the task of automated selection of safe and effective training programs. Unlike many analogues, the system prioritizes medical restrictions, enhancing the safety of the training process. The conducted techno-economic analysis confirmed the project’s investment attractiveness, calculating a payback period for development costs of 1.60 years. Keywords: decision support system (DSS), personalized planning, gym training, hybrid method, analytic hierarchy process, UML. References. 1. Нейромережі для спорту і фітнесу: Використання ШІ для персоналізації тренувань, аналізу результатів і прогнозування спортивних досягнень - Нейросенсей. Нейросенсей. URL: https://neurosensey.com/nejromerezhi-dlya-sportu-i-fitnesu-vykorystannya-shi-dlya-personalizacziyi-trenuvan-analizu-rezultativ-i-prognozuvannya-sportyvnyh-dosyagnen (дата звернення: 24.10.2025). 2. Calorie tracker & BMR calculator to reach your goals | MyFitnessPal. URL: https://www.myfitnesspal.com (date of access: 24.10.2025). 3. Intensive workouts & individual training plans | FREELETICS. URL: https://www.freeletics.com (date of access: 24.10.2025). 4. Глибоке навчання – Вікіпедія. Вікіпедія. URL: https://uk.wikipedia.org/wiki/Глибоке_навчання (дата звернення: 25.10.2025). 5. Школа фітнес інструктора - Технології для носіння - лідер списку фітнес трендів. Школа фітнес інструктора. URL: https://www.fitnesservice.com.ua/статті-та-фото/технології-для-носіння-лідер-списку-фітнес-трендів (дата звернення: 25.10.2025). 6. Гайд з розробки віддаленого моніторингу пацієнтів | Stfalcon. Logistics & Transportation Software Development Agency | Stfalcon. URL: https://stfalcon.com/uk/blog/post/health-monitoring-system (дата звернення: 25.10.2025). 7. Мобільні застосунки фітнес-тренера для роботи з клієнтами. Оранжева книга тренера. URL: https://orangebook.apollo.online/mobilni-zastosunki-fitnies-trieniera-dlia-roboti-z-kliientami/ (дата звернення: 25.10.2025). 8. Теленик С. Ф., Суржик Л. М. Системи підтримки прийняття рішень: навч. посіб. Київ: НТУУ «КПІ», 2010. 9. Литвин В. В. Технології проектування інформаційних систем: навч. посіб. Львів: Видавництво Львівської політехніки, 2014. 10. Fowler M. UML Distilled: A Brief Guide to the Standard Object Modeling Language. 3rd ed. Boston: Addison-Wesley, 2003.