Real-Time Collection Method of Athletes’ Abnormal Training Data Based on Machine Learning

Real-time collection of athletes’ abnormal training data can improve the effect athletes. This paper studies real-time method based on machine learning. The main motivation this is to collect in time, which help evaluate and effect. Four sensor nodes are arranged upper lower limbs athletes angular velocity, acceleration, magnetic field strength state. sent transmission base station through wireless sensors, transmits processing terminal. terminal calculates difference between sample values each obtain dispersion sensor. features dimension a time domain frequency obtained by using degree construct 32-dimensional feature vectors, extracted vectors input into hidden Markov model. forward algorithm used probability final observation sequence, so as realize data. experimental results show that accuracy recall rate collected higher than 98%, requires less time.