kalika12 / customer-transactionprediction

Customer-TransactionPrediction

The objective of this project is to analyze the datasets results of the bank to identify which customer will make transactions in future ,irrespective of the amount of money transacted .

Dataset had an imbalance data with the majority of negative result distribution. As the bank's data scientists, we decided to focus on minimizing both False Negative Class and False positive class.

Evaluation metric that we will use is F1 Score. 1) In False positive class where we predicted clients that will not made a transaction, turns out they actually made a transaction . 2) In False negative class where we predicted clients that will made a transaction, turns out they actually not made a transaction .