Level: Intermediate
【Presenter】
Seiichi Yasui, Lecturer, Faculty of Science and Engineering, Department of Business Engineering, Tokyo University of Science
[Announcement summary]
Before performing data analysis, it is necessary to create a data set for analysis by extracting necessary data, changing / shaping the correspondence between variables, changing variables / categorizing / recategorizing, etc. JMP provides computational functions required for row and column extraction, database operations such as joins, and change of variables. Data sets for analysis can be easily created using these functions. However, it is necessary for the analyst to instruct operation instructions such as setting the extraction target and change of variables, and when the operation instructions become complicated, there is a high possibility that the intended result has not been obtained. For example, when setting a range in data extraction or categorizing by an if statement, the more complicated the "and" and "or" rules are, the more likely it is that the desired data set for analysis has not been obtained. Therefore, it is necessary to mechanically check whether the data set for analysis matches the one intended by the analyst.
The quality of the data set for analysis can be confirmed by the statistical method of JMP. The method to find the maximum and minimum values of a variable is the simplest, but the "distribution of one variable" and "relationship of two variables" are also strong, and the contribution rate of 1 is evidence in the "relationship of two variables". Is.
In this presentation, we will report an example of confirming the quality of the analysis dataset for large-scale data.
[Presenter Profile]
Lecturer, Department of Business Engineering, Faculty of Science and Technology, Tokyo University of Science. Senior Researcher, Department of Chemical Systems Engineering, The University of Tokyo. His research specialty is statistical quality control. I am mainly researching statistical analysis methods necessary for quality control, but I am also applying statistical quality control to fire prevention construction, fire phenomena, medical care and nursing care, and modeling large-scale data using JMP. The information behind it is extracted and fed back to the unique field to be researched.
