Research On Individual Sports Health Evaluation Method Based On Restricted Boltzmann Machine

Health is the basis for survival and development of human.With the continuous progress of society and the continuous development of science,people's living standards have been greatly improved,but all kinds of diseases and sub-health are coming.In order to improve the quality of life and enjoy a more comfortable life,people are paying more and more attention to physical health.They are also pursuing scientific fitness exercise and gradually realizing the importance of moderate fitness for health.Restricted Boltzmann machine is a generative stochastic neural network.which is composed of visible and hidden units,and it expresses the correlation between the units through the weight value.There is no connection in the visible layer or hidden layer,and all layers are fully connected.Restricted Boltzmann machines include many algorithms,such as Gibbs sampling,gradient descent,and contrastive divergence.Restricted Boltzmann machine has received extensive attention in recent years.As a practical application,the algorithm of restricted Boltzmann machine shows its superior learning performance in many aspects,and occupies an important position in the learning of neural network.This paper firstly introduces the background of sports health,analyzes the applicability and research significance of restricted Boltzmann machine in the field of sports health,and expounds the requirements of modern urban life for scientific and effective fitness exercise methods.Then a brief introduction is given to the restricted Boltzmann machine,including introduction of some basic models and introduction of the traditional learning algorithm,and detailed description of the setting of training parameters.Therefore,based on the previous work,this paper studies the individual motion health assessment method based on restricted Boltzmann machine.The main tasks are as follows:(1)Aiming at the defects and deficiencies of the traditional restricted Boltzmann machine,this paper improves the traditional restricted Boltzmann machine by adding one impulse item to the connection weights,and the unit bias and the hidden layer unit bias,and establishes a model of exercise intensity measurement method based on restricted Boltzmann machine,take common physiological indicators such as heart rate,lactic acid and relative oxygen uptake as input,measure the individual's exercise intensity,Through the comparison and analysis of the traditional and improved Restricted Boltzmann machine on the measurement of motion intensity,the corresponding conclusions are drawn and reasonable suggestions are given in order to ensure the safety of individual fitness exercise.(2)On the basis of the improved Restricted Boltzmann machine,we stacked several constrained Boltzmann machines to form a deep belief network,and a method for predicting the optimal amount of motion based on restricted Boltzmann machine is proposed.Through the analysis of the individual's physical condition before and after exercise,it is predicted.The best amount of exercise improves the efficiency of individual fitness exercise.Finally,this paper will carry out a large number of experiments on the above evaluation methods,and prove the validity of the above two assessment methods through the experimental data,and will also verify the reliability and versatility of the exercise health assessment methods in this paper.In short,the purpose of this paper is to use the learning algorithm of restricted Boltzmann machine to propose an effective health assessment method and model,real-time monitoring of its state of motion,and some reasonable suggestions are given to ensure the safety and efficiency of individual fitness exercise for a long time,so that people can carry out fitness exercise in order to meet the needs of modern healthy life.