Regularity of Pediatric Common Diseases Based on Data Mining

  • Rui Wang


Chinese medicine is the quintessence of China, but it has not been well developed and utilized. The raw data collected by Chinese medicine hospitals about common diseases in children is increasing. Many patients’ drug information and raw data such as various cases are stored. These increasing drug data implies a lot of great value. How to acquire and mine this high-value knowledge is the key to today’s research. As a very effective method of knowledge acquisition, data mining extracts high-value information from the database of common diseases in children. Then, by analyzing and evaluating these data, some potential data patterns can be found and scientific medical judgment and treatment can be provided. In this paper, Chinese medicine is used to treat cough in common pediatric diseases as an example. This paper first analyzes the characteristics of cough medication information in children, and cleans, integrates, transforms and other such information to standardize the data. Then, a method for analyzing the information of pediatric cough medication is analyzed, which shows the possibility of data mining in the application of drug law. The corresponding model is established by using the characteristics of the drug type and medicinal properties in the medication information. Using k-means algorithm to be simple, efficient and easy to implement, predict the use probability of different medicinal tastes, different medicinal properties, etc., and use the hierarchical clustering algorithm to predict the shortcomings of the k-means algorithm. Then, the algorithm is improved on the defects and arguments of the traditional k-means algorithm and hierarchical clustering algorithm. Through experimental proof and comparison with each other, the hierarchical clustering improvement algorithm can classify the cases more accurately, thus determining the importance of various properties of drugs for drug use.