FinTech Project B21 - Fidelity - Procurement Categorization of the Current Spend Trends

Reneson, Olivia; Zeng, Ziqian; Deratzou, Luke

Student Work

FinTech Project B21 - Fidelity - Procurement Categorization of the Current Spend Trends

Public

Fidelity utilizes spend procurement software to make a multitude of financial decisions. Before the data can be available within the software, the data needs to be classified. Fidelity uses machine learning models to classify spend data and would like the models to be more accurate and efficient to leverage spend data. Miscategorized spending distorts documentation and reduces the value of the spend data. To enhance the classification process, Fidelity has asked our team to improve the current models by using Python multithreading on Amazon EMR and introducing models with lower run time and higher accuracy. Our team met Fidelity’s needs by introducing two new, faster models (K Nearest Neighbors and Decision Trees) as well as fine-tuning pre-existing models to increase accuracy.

This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator

Subject

Publisher