Research On Algorithms And Applications Of Big Data Generation Based On Statistical Learning

Big data analysis and handling technology is one of hotspot issues in the computer field.However,big data are oftentimes accompanied by specific business information and trade secrets,and it is difficult for researchers to obtain effective big data.Therefore,how to generate simulation big data technology based on limited real data has become a problem that needs to be studied and solved in the academic community.Specifically,a large amount of training data is required to establish a machine learning model for big data analysis,and a large amount of test cases are required for a big data processing system.In response to the above technical requirements,this article discusses two large data set generation algorithms and application technologies.Including:discrete cascade and large data generation algorithm and the Beijing traffic congestion prediction application cases.algorithm is proposed based on statistical learning theory and boundary diffusion techniques iterative algorithm for large data generated by diffusion boundary-setting generation data set,continually generate simulated data sets within the boundary until the big data applications to meet the requirements.research using real traffic data collection Beijing car GPS as the experimental initial data set to generate a large data set on this basis.The generation of large data sets for analysis of traffic congestion in Beijing,testing the effectiveness and availability of the algorithm.The similarity between the generated big data set and the real big data set was evaluated,and the similarity reached 70%.