We plan to work with a set of data that contains 10,000 customers from banks that consists their age, salary, credit card limits, current banking status, and other information. There are 18 features overall. We aim to use this dataset to determine and find out the characteristics of a potentially “turned customer” so that banks can come to them first and offer more services or better deals in order to turn them back.The data is originally collected from an even larger data set from https://leaps.analyttica.com/home. It consists of clients information from a certain banks. It is collected due to a concern from a bank manager that he found out more and more customers start to leave the bank, and he thought it would be great if they could find out these customers before they “turnd” and offer them better services to keep them.Yes, we are still working on it. What characteristics do “truned customers” have in common? Or predict what type of customer is likely to be turned? Some of the demographic variable and other data are categorical data, including Educational Level and Gender, etc. These variable are valuable to our analys, which lead us to think about how to sucessfully indicate the absence or presence of some categorical effect that may be expected to shift the overall outcome. We intend to use dummy variable, numeric variable that represents categorical data, for this intention.Also, the data we have may have limitations and could be biased. It lacks variability and may not see the whole picture of customer churning.