Research On The Correlation Between Zhongyun Theory And Relative Humidity Based On Data Mining

The theory of five evolutive phases and six climatic factors(the theory of Yunqi for short)is an important basic theory of traditional Chinese medicine.It is a theory in ancient China that studies the laws of time and climate change,as well as its impact on the human body.Zhongyun is a part of the theory of Yunqi,which can be used to deduce the regularity of climate,phenology and disease of the whole year.Understanding the background of the emergence and development of the theory of Yunqi from the perspective of meteorology is the research trend of the basic theory of traditional Chinese medicine in recent decades,but there are still deficiencies in the scientific nature,normalization and result evaluation of research methods.In this study,we advance clustering as a new verification method of relation between Yunqi and meteorology,which is widely used in the field of machine learning and data mining.Taking the correlation between national relative humidity dataset and Suiyun as a research case,we refer to the characteristics of meteorology and deduction of Yunqi to finish the specific application and applicability improvement.The paper will be discussed from the following aspects.1.Using statistics and calculation methods to verify the correlation between relative humidity and Zhongyun.Based on the 136 meteorological stations which have 60 years complete relative humidity data in china,the paper analyzes the correlation between the relative humidity data and Zhongyun from the national and regional perspectives.From the national perspective,it is found that the relative humidity data has a very high coincidence rate with Zhongyun based on the loose pattern.From the regional perspective,it is found that the actual change of relative humidity is not completely consistent with the deduction of Zhongyun,but there is still a relatively stable correlation between them.It shows that the theory of Yunqi is reasonable and reliable.2.Weakening the research method of using mean as the conventional judgment standard and using clustering method to find the implicit correlation between relative humidity and Zhongyun.This study is an extension of the regional perspective research.It changes the method of administrative zoning and discusses Zhongyun based on the areas division,which is stemed from relative humidity in different Heavenly Stems.Based on the initial data of relative humidity,the study constructs two kinds of features,and respectively uses k-means algorithm to cluster the sites.It is found that there are two similar optimal clusters in two clustering results,which are almost in perfect agreement with the Zhongyun deduction,and the sites collection of two optimal clusters are similar.The results of clustering experiment show that the region distribution of three clusters brings enlightening conclusion about the birthplace of Yunqi,global deduction and spatiotemporal structure,etc.3.Using association rule mining to analyze the correlation between relative humidity and Zhongyun.In the study,Apriori algorithm is used to mine association rules,and 24 strong rules based on the relative humidity characteristics of the ten heavenly stems between stations are found and analyzed.The results show that,the relative humidity changes in some Heavenly Stems are stable and consistent with the Yunqi theory,which are the positive factors and potential reason why most sites’Yunqi inference has a high coincidence rate.On the other hand,not all Yunqi inference can form strong association rules,so the guiding significance of the theory of Zhongyun is limited.