The Applied Research Of Big Data On International Clinical Trial Registration

With the advent of the era of big data,big data-related technologies are also increasingly developing and gradually maturing.As one of the most valuable industry big data,medical big data has been deeply researched in recent years.This article is aimed at the limitations of traditional clinical trial registration schemes in medicine,and adopts a method of combining big data technology with clinical trial registration data in medicine.Provide supplementary reference for trial participants,sponsors and subjects,and improve the efficiency of program design.This article first crawls the data of the clinical trial registration center through the Scrapy framework system.The crawler framework was expanded to distribute the original crawler tasks to several small queues and distribute them into several crawler tasks.The parallel processing execution method was adopted to improve data acquisition efficiency.Clean and preprocess the acquired data,merge or delete the attributes of the data according to certain rules,and delete the repeated clinical trial registration to form the data for analysis.Then through the ARIMA model and neural network LSTM model in big data to model the registration of clinical trials.The ARIMA(p=1,d=1,d=1)model,the LSTM model with a step size of 1 and the LSTM model with a step size of 12 were constructed separately.Use the constructed model to predict and compare the total registration of clinical trials in order to grasp the trend and fluctuation status of clinical trial data.Secondly,Naive Bayes algorithm,K nearest neighbor algorithm,logistic regression algorithm,SGD stochastic gradient descent algorithm,(SVM)support vector machine algorithm,decision tree algorithm and random forest algorithm are used to predict the sample size.And try to improve the random forest algorithm and exclude the number of trees in the forest based on the decision tree classification accuracy,so that the random forest algorithm has a higher accuracy in predicting the sample size.Finally,the big data visualization technologywas used to analyze the cases of breast cancer and anti-tumor end point indicators,and to visualize the trial design schemes such as the regional distribution of clinical trial registration of breast cancer and anti-tumor end point indicators,blind selection,and randomized control methods.Analyze and summarize the design rules of breast cancer and advanced breast cancer clinical trial registration,and also summarize the regularity of anti-tumor endpoint indicators,such as regular measurement of tumor and period distribution,to provide participants and designers of clinical trial registration with more intuitive and comprehensive design information.