This project applies statistical analysis, forecasting, regression, optimization, and simulation techniques using R to solve business problems in retail, customer loyalty, staffing, and loan approvals. It focuses on deriving insights to support decision-making and improve performance.
- Top Products: Identified best-selling products by region and season using revenue and ranking.
- Hypothesis Testing: Tested if average revenue differs across regions using Kruskal-Wallis.
- Regression: Built simple and multiple models to predict revenue; evaluated and compared results.
- Customer Loyalty: Used logistic regression to analyze factors affecting loyalty and validated model fit.
- Time Series Forecasting: Applied SMA, SES, and Holt’s method to forecast sales; compared accuracy metrics.
- Staff Allocation: Optimized staffing to minimize costs while meeting operational needs using linear programming.
- Loan Approval: Built a logistic tree and performed Monte Carlo simulations to analyze risk and approval rates.
- Language: R
- Libraries: dplyr, ggplot2, readxl, forecast, car, rpart, lpSolve, pscl, etc.
- Methods: Statistical tests, regression, forecasting, optimization, simulation, visualization.
- Pricing and discounts vary by region and season.
- Customer behavior impacts loyalty significantly.
- Forecasting helps plan inventory and staffing.
- Simulations assist in balancing risk and approval thresholds.