Research And Implementation Of Wideband High Resolution Frequency Synthesizer

The era of big data has come.Data processing,analysis,and discovering new learning models have become indispensable jobs in every industry.In the environment of big data,the application of data mining technologies,such as time series analysis,association rule mining,clustering analysis,classification and prediction methods of machine learning is ubiquitous in every walk of life.Chronic Non-communicable Diseases(CNDs)have a huge impact on global human health and a serious burden of disease.In China,CNDs have become the dominant disease among urban and rural residents.In recent years,studies on the risk factors of CNDs have also increased gradually.In this paper,time series analysis,cluster analysis,association rule mining and machine learning techniques in data mining technology are effectively used to analyze and process inpatient medical data of patients for 27 CNDs such as hypertension and diabetes,so as to provide scientific basis for the prevention and treatment of CNDs.This paper focuses on the study of the health impact of air pollution on inpatients with CNDs,the association mining of CND complications,and re-admission risk prediction of hospitalized patients with CNDs.The research work is divided into the following three parts:(1)Correlation analysis of air pollutants and CNDs based on generalized additive model.The generalized additive model(GAM)in time series analysis was used to study the acute health effects of air pollutants on patients with CNDs,and stratified experiments such as age,sex and season were used as sensitivity analysis to ensure the stability of experimental results.(2)Association rule mining for chronic diseases and their complications based on hierarchical clustering.Based on cluster analysis and association rule mining,CNDs complication mining was carried out in this paper.Firstly,27 CNDs were clustered using three clustering methods(k-mean ++,average-linkage method and error sum of method).FP-growth algorithm was used to obtain the complication relationships between CNDs in each cluster,after that,the disease network was established.Then,comorbidity combinations were found in the whole disease group for the three major CNDs.(3)Re-admission risk prediction of hospitalized patients with CNDs based onmachine learning.In this paper,the basic readmission risk prediction model was established based on logistic regression and three integrated learning models including random forest,Gradient Boosting Decision Tree and light GBM.Finally,a hybrid model based on compressed storage is proposed,which combines tree model and logistic regression model to improve the prediction performance.