Research On Intelligent Data Analysis Model And Algorithmi For Cardiovascular Diseases

With the development of economy and the change of people’s lifes-tyle,cardiovascular diseases are one of the diseases that endanger human health,and the patient mortality rate is high.Through deep learning,data mining and other intelligent methods,judging and diagnosing the patient’s condition based on the heart beat,blood pressure,blood lipids and other vital indicators are important for the early prevention and treat-ment of the patient’s condition.Based on the characteristics of medical bi-g data and the deficiency of existing intelligent models,the paper propos-es and establishes a model of cardiovascular disease discovery and diagn-osis.The main research contents of the paper are summarized as follows:(1)In the process of disease discovery and diagnosis,the nonlinearity,time-variation and complexity of the heart beat data are discussed.The paper superposes the sparse automatic encoder and the process neural network stack to build a deep process neural network model(DPNN).The DPNN model maintains the diversity of sample features and effectively improves the ability to extract signal structure features and the discrimination of features of different samples.At the same time,the traditional deep neural network is expanded into a time domain in the information processing mechanism,and the time-varying signal is directly classified and processed.The actual analysis and treatment of cardiovascular disease predictions have yielded good results.(2)For the intelligent diagnosis of cardiovascular diseases based on various vital signs of the human body(such as blood oxygenation,resting blood pressure,etc.),this paper has established a dynamic clustering algorithm based on fuzzy clustering.The model uses an iterative approach to combine the fuzzy clustering algorithm with the DBSCAN algorithm,eliminating the dependency of the model on the initial setting of the cluster center and the number of clustering categories.It also integrates the advantages of the DBSCAN algorithm and improves the effectiveness of clustering algorithms.(3)For the intelligent diagnosis of cardiovascular diseases,the data in the patient data set is incrementally changed,and the patient data is generated at a high speed.The paper,on the other hand,combines the grid pruning algorithm and the LOF algorithm to propose a dynamic clustering model for rapid diagnosis and treatment.This model combines the advantages of the speed of the grid algorithm has nothing to do with the amount of data,and the characteristics of the LOF algorithm for incremental database analysis.It greatly shortens the time of disease discovery and diagnosis and realizes the analysis and processing of dynamic databases.(4)For the diagnosis process of cardiovascular diseases,the distribution of medical data in various remote nodes and the clustering rules obtained by the traditional clustering algorithm can not be applied to the new data sets.The paper establishes a distributed-based discovery and diagnosis model for cardiovascular abnormalities.Based on the weighted integration weights,this model uses a deep neural network to integrate each sharing model to obtain a DDM model.This model improves the accuracy,adaptability,applicability,and privacy of cardiovascular abnormality discovery and diagnosis mining algorithms.