Bank Telemarketing Analysis: Implemented Logistic Regression to predict the outcome of the telemarketing campaigns based on the characteristics of the clients and the calls with 91.14% Accuracy.
Telemarketing is a method of selling products and services over the phone to customers. It has always been a controversial approach. On one hand, it is easy to directly reach out to customers and also cheaper than other marketing methods. On the other hand, it has bad reputations of damaging the company's image and some of the startup costs are very expensive.
In this project, we are looking into how other factors can affect the outcome of telemarketing campaigns for a specific institution, Portuguese retail bank, and make prediction based on our model. The main focus of this project is incredibly interesting since we typically feel annoyed by telemarketing.
Our main question is what is likely to be the outcome of the telemarketing campaigns based on the characteristics of the clients and the calls. Our goal is to predict if the client will subscribe (yes/no) to a term deposit (variable y).
Note: A term deposit is a type of deposit account held at a financial institution where money is locked up for some set period of time
• Dataset name: Bank Tele-Marketing Data Set
• Original Source of the Dataset: [Moro et al., 2014] S. Moro, P. Cortez and P. Rita. A Data-Driven Approach to Predict the Success of Bank Telemarketing. Decision Support Systems, Elsevier, 62:22-31, June 2014
• We retrieve the data from UCI Machine Learning Repository. The data is accessible here.
• The data is collected from several telemarketing campaigns in which the Portuguese bank attempted to target customers through phone calls to sell long-term deposits.
• The dataset includes both the phone calls of which the bank executed and the phone calls of which clients contacted the help center.
• Each observation includes the outcome, whether or not the target customers subscribed the term deposit, and the characteristics of the customers and the phone calls themselves.
• Data: bank-additional-full.csv
• R markdown with codes and interpretation: bankfinal.Rmd. The project uses R 4.1.2
• A detailed report: bank-project-report.html
• A powerpoint for presentation purpose: bank-project-presentation.pdf
[Moro et al., 2014] S. Moro, P. Cortez and P. Rita. A Data-Driven Approach to Predict the Success of Bank Telemarketing. Decision Support Systems, Elsevier, 62:22-31, June 2014