Case Study
The client, an online retail marketplace company, was keen to extract valuable insight from survey data relating to attributes that affect customers’ willingness to purchase ads.
Solution
- Gathered business intelligence through statistical techniques to derive key variables that drive customer behaviour
- Employed Bivariate Analysis to explore the relationship between predictor and dependent variables to check for associations as well as their magnitude
- Binned dependent variable into categories to reduce noise due to minor data inconsistencies
- Developed hierarchical rules which are easy to interpret and intuitive to describe using tree-based models like CART & Random Forest algorithms
Value Addition
Utilisation of tree based models to optimize pricing decisions by taking into account various attributes of customers, ads and their interaction
→ Related Service Offering: Private Equity Analytics as a Service