Projects

As a requirement for my Master’s degree, I took part in a practicum project with a local client, Louisiana Lottery Corporation. For this project, I worked on a team with 4 of my MSA classmates and served as an analytics consultant. We were asked to identifying significant indicators that could impact the sales of scratch-off games. As MSA consultants, we analyzed the variables associated with instant play game sales to identify possible opportunities that would allow Louisiana Lottery to potentially increase sales through improved re-stock and launch scheduling of instant play games. My team and I conducted in-depth data extraction from all of LA Lottery’s scratch-off working papers, explored regional trends, and created a forecasting model at the individual retailer level to optimize product scheduling. The forecast results were to predict weekly sales by quantity and price point for each retailer. Backtesting was performed on the data in order to calculate certain accuracy metrics, such as the Mean Absolute Percentage Error (MAPE) which was about 30%. The forecasting model was created with Amazon Forecast from AWS. The forecast model used was a Neural Network, specifically CNN-QR, Convolutional Neural Network - Quantile Regression. The forecast results were exported and ingested into Microsoft Power BI for visualization. Knowing LA Lottery currently uses Power BI, we felt it was logical to view the data this way. A large portion of my role consisted of data exploration using Tableau. Visualizations of game color distribution, key word frequencies, and regional trends inspired our clients to view their traditional re-stock and launch processes in a new way.