In order to solve the problem that the abnormal heart rate data of low amplitude athletes can not be forewarned, and to achieve a comprehensive early warning of abnormal data objects, the online early warning of abnormal heart rate data of athletes considering individual differences is studied. According to the sampling results of classifier performance indicators, the individual differences of athletes’ heart rate are analyzed, and then based on this, the optimal feature subset of abnormal heart rate data is defined to optimize the characteristics of abnormal heart rate data considering individual differences. The QRS wave group of abnormal heart rate data is detected, and the online pre-warning process is improved by optimizing the SVM pre-warning parameters, so as to complete the design of online pre-warning method for abnormal heart rate data of athletes considering individual differences. The experimental results show that the application of the above methods can ensure that the early warning accuracy of low amplitude athletes’ abnormal heart rate data is higher than 90%, which meets the practical application needs of comprehensive early warning of all abnormal data objects.