Predict Credit Card Consumption
- Understanding the consumption pattern for credit cards at an individual consumer level is important for customer relationship management. This understanding allows banks to customize for consumers and make strategic marketing plans. Thus it is imperative to study the relationship between the characteristics of the consumers and their consumption patterns. XYZ Bank has given a sample of their customers, along with their details like age, gender and other demographics. Also shared are information on liabilities, assets and history of transactions with the bank for each customer. In addition to the above, data has been provided for a particular set of customers’ credit card spend in the previous 3 months (April, May & June) and their expected average spend in the coming 3 months (July, August & September). Predict the average spend for a different set of customers in the test set for the coming 3 months.
- The challenge is to understand the consumption pattern for credit cards at an individual consumer level.Datasets contains all consumer information.Thus it is challenging to study the relationship between the characteristics of the consumers and their consumption pattern .
- iVentura Machine Learning Platform was used for building the solution. iVentura provides the complete ecosystem for data scientists to build models without worrying about the underlying Infra & Security. Either for a team or an individual data scientist, iVentura is ideally suited as a platform of choice.
- To deal with the above problem statement ,datasets needs to be analysed and evaluated with metrics to acquire best outcome. Here we go:
- 1) Input Datasets is zipped file.Zip file is extracted and read traon.csv file
- 2) For Understanding the consumption pattern for credit cards at an individual consumer level ,Exploratory Data Analysis is performed on datasets
- 3) KNN algorithm can get best consumption pattern of consumer’s information. Thus ,Best parameter “k” is taken with lowest RMSE value.
- 4) Train the model using KNN with best ‘k’ and predict consumption details.
- 5) Save result pickle
- 6) Deployment & Visualization