This paper proposes an intelligent algorithm-supported design framework for athletes’ physical fitness improvement paths, aiming to optimize the training program by grouping data mining techniques and clustering algorithms. Eight training groups of athletes from different regions of city A were selected as research objects, and the Apriori algorithm was used for in-depth correlation analysis of physical fitness data. K-means clustering algorithm was introduced to determine the grouping for the athletes’ physical fitness data, and personalized training programs were formulated relying on the characteristics of each group. Ten athletes were randomly selected for investigation, the average number of times the heart rate of the ten athletes fell in the effective interval without program guidance was 8.2 times, and the average value of the number of times the heart rate of the ten athletes fell in the safe and effective interval under the personalized training program was 9.8 times, which is 19.51% more than that of the training without program guidance, which indicates that the intelligent algorithm provides a scientific and precise path design for the improvement of athletes’ physical fitness.
