Research On Application Of Data Mining In Laboratory Information Management System

The laboratory information management system used by the CDC is in line with international norms.It can ensure that the testing data of the internal control center of the disease control center conforms to the national ISO/IEC 17025 standard and can better ensure the authenticity of the data and data.Structural integrity.This article takes the laboratory information management system as a platform and uses data mining technology to analyze the correlations among the testing samples.The specific work is as follows:1.Select the relevant data of 20 kinds of samples of a disease control center in 2016 as experimental research objects,conduct data modeling on external factors such as detection methods,instruments,and health standards that affect the detection results,and adopt many evaluation indicators(such as accuracy Relative root-mean-square error,ROC area,etc.)Compare the quality of various classification models(mainly using decision tree,OneR and Naive Bayes algorithm);2.Analyze the internal factors of the samples that affect the test results and determine whether there is a strong correlation between the samples within the sample.3.Using the clustering K-Means algorithm,the unqualified samples and items are brought together to analyze the common features of the unqualified samples.The research results show that the data model established by using the decision tree C4.5 algorithm has the best prediction effect,the OneR classifier takes the second place,and the naive Bayes classifier has a poor prediction effect.By correlating the Apriori algorithm,it was found that there are strong correlations between certain indicators inside the test sample,such as the detection of drinking water,odor and taste,turbidity,color,and visible substances.Through the K-Means cluster analysis,the month of unqualified sample collection,sampling department,sample type,and category of detection items can be found.This study applies data mining to the disease control center laboratory information management system,which plays a good role in improving the accuracy of the disease control center for sample detection and disease prevention,and transforms sample detection big data into disease.Prevention and diagnosis provide a reference model.