Project Overview
A telecommunication business recently began a marketing campaign to encourage clients to sign up for their new subscrip- tion plan. The organization is looking for help in developing a complete understanding of its consumers and identifying the client group that are most responsive to marketing initiatives.
The project aims to develop statistical models that can predict the success of a marketing campaign for a given customer. Based on developed models, customers will be segmented based on responsiveness likelihood to illustrate the difference between groups and deliver optimization strategies for business’s marketing campaigns.
Responsibilities
Applied various EDA techniques, including visualization, univariate and bivariate analysis, etc. to draw comprehensive insights of customers and external contexts of the given dataset.
Applied hypothesis testing to verify generated hypothesis of variable relationships.
Developed various classification models, including XGBoost and TPE hyper-params optimization to extract insights from different customer segments.