BIG PICTURE
Determine what factors affect client churning OR the characteristics of attrited customers have in common.
We plan to do correlation analysis between each variable in a big picture. The planned format is a correlation coefficient interactive form. After that, we can see which variables have stronger correlation with each other, helping us in later modeling part to decide the interaction term.
Also, we decide to use the interactive ways to help people know our dataset, especially the categorical variables by displaying proportions of each category.
INTERACTIVE Download flexdashboard package Present directly what factors affect client churning
ANALYSIS Convert our categorical variables and use numbers to represent each different values, for example, in attrition vs. education level, Parameters are: 1-College 2-Post-Graduate 3-Graduate 4-High School 5- 6-Uneducated 7-Unknown. Conduct linear regression models based on different factors like Age, Gender vs Attrited Customers, or Education level and age vs. Attrited Customers etc. Just to find out relationships between attrited customers and other potential influential customers. Use ggplot to graph