Date of Graduation

5-2026

Document Type

Thesis

Degree Name

Bachelor of Science in Data Science

Degree Level

Undergraduate

Department

Data Science

Advisor/Mentor

Karl Schubert

Committee Member

Kelly Sullivan

Second Committee Member

Eric Specking

Third Committee Member

Zach Bonds

Abstract

The purpose of this research is to implement an OpenAI Reinforced Learning prescription-giving model for improving sales on a week-by-week basis. The data used comes from a segment of High Impact Analytics’s sales data that has been anonymized for proprietary reasons. The features among the data include inventory numbers, shipments in transit, total quantity and dollars of products sold each week for the past 2 years, all aggregated at the store-item-week level. In order to build this model, Tigramite, a causal discovery model combined with prediction models XGBoost, Linear Regression, Ridge Regression, Lasso Regression, Scikit-learn’s MLP, and Keras’s Neural Model were implemented (Anysphere, 2026; Brockman et al., 2016; Chen & Guestrin, 2016; Hunter, 2007; McKinney, 2010; Pedregosa et al., 2011; Python Software Foundation, 2024; Raffin et al., 2021; Runge et al., 2019; Towers et al., 2024). First, the Tigramite causal model was trained on the data aggregated on a time-basis in order to find time-related relationships between variables, which resulted in finding that store on order quantity, store on hand retail, max shelf quantity, and customer returns have causal relationships with next week’s sales. Next, the predictive models were trained on the unaggregated data with the goal of predicting next week’s sales. This orchestra of models then served as the foundation for the OpenAI Reinforced Learning model. This model would create a virtual environment and learn how next week’s sales changes after altering the features that have causal relationships with next week’s sales, resulting in a model that can be fed one week’s sales data and prescribe the best action for raising next week’s sales.

Keywords

machine learning; reinforced learning; causal discovery; tigramite; causal relationships; PCMCI

Download

Included in

Data Science Commons

Share

COinS