Research On Blood Pressure Detection Method Based On Big Data Analysis

With the improvement of living standards,the number of patients with high blood pressure continues to climb,and for people who are recovering from a serious illness,hypotension can cause insomnia or syncope,so attention to high and low blood pressure is very important.Today,big data technology is developing rapidly.Using it to accurately monitor and analyze blood pressure is of great significance to solve the problems caused by high and low blood pressure.In view of the problems of non-standard blood pressure measurement methods and low prediction accuracy,this paper proposes a new scheme based on big data analysis and using pulse and blood oxygen to predict blood pressure,focusing on how to combine grid search and cross-validation respectively Optimize the parameters of the support vector machine and the artificial neural network model to improve the analytical performance of the algorithm model,so as to better achieve the analysis and prediction of high and low blood pressure.The main research work is as follows:(1)Combine the MIMIC database and sensors to obtain data,and use Hbase and other big data platforms to process the data.(2)Build Hadoop and Spark big data platforms,build support vector machines,artificial neural networks,Logistic regression and Lasso regression models on this basis to analyze and predict blood pressure,and use accuracy and root mean square error for evaluation.(3)Introduce grid search into the support vector machine and artificial neural network model for blood pressure analysis and prediction,and use cross-validation to find the parameter pairs that make the model performance optimal.(4)Construct the support vector machine and artificial neural network model under the optimal parameter pair,and input the pulse and blood oxygen data into the model for evaluation by using the accuracy and root mean square error.Compare the evaluation value with the result value obtained in step(2).The results show that the prediction accuracy of the support vector machine after parameter optimization for high and low pressure is about 71.39% and 81.69%,and the root-mean-square error is about 0.5349 and 0.4279,which is obviously better than the other regression algorithms.The effectiveness of the improved support vector regression algorithm based on grid search is verified in this paper.