Insight the customers’ transactions behavior, some interesting thing I observation. The avarage transaction volumn by each customer is 35 times, and people like to purchase thing in Wed. and Fri. Depond on my observation, I predicted the customers’ annual salary for helping banking can segment customers into income brackets. In this project, I use python and Google Map Api to draw the transaction location. To improve my model accuracy, tring seveary models to classificate customers’ transactions behavior.