Aim of this code is to find the most accurate and precise model to predict, which clients (test data) will stay and which are hesitant and might plan to leave the company.
Dataset has following attributes:
Rownumber: Unique ID for every row
CustomerID: Unique ID for every client
Surname: Client's surname
CreditScore: Client's credit score
Geography: Country of client's origin
Gender: Client's gender
Age: Client's age
Tenure: Number of years for which the client has been with the bank
Balance: Client's balance on account
NumOfProducts: Number of client's products
HasCrCard: Flag whether client has credit card or not
IsActiveMember: Flag whether client is active member of bank or not
EstimatedSalary: Client's annual estimated salary in euros
Exited: Target variable, flag, whether client left the bank or not
Used random over sampler since target variable was highly non-homogeneous.