domains. Variation in sourcing, processing, and methodology can result in different values across vendors and markets for the same dataset.
A 200% price movement could be the result of a bad tick, or it could be a real earnings shock. Detecting anomalies is only part of the battle; the bigger challenge is validating those anomalies across millions of records to separate actual issues from legitimate market moves.
That was the challenge presented to Johns Hopkins Carey Business School's Analytics Consulting Project, where QUODD's Data Quality Manager @Ritu Ghosh worked with a team of Business Analytics and AI students to analyze historical pricing for valid data anomalies.
We are honored to have sponsored and participated in this project, and we're excited to share with Carey Business School the case study detailing the creative solutions the students came up with in just eight short weeks.
Despite the complexity of the challenge, in just eight short weeks, the students built ingenious solutions in Python to flag, validate, and, most importantly, categorize pricing anomalies, data gaps, and missing corporate action adjustments by severity in QUODD's historical pricing data. The result: production-ready code that our data analysts and engineers can easily put to use.
Congratulations to Jialu Chi, Xinyue Xing, Yumeng Jiang, Yuanheng Fang, and Qiyue Yao. Thank you also to their faculty coaches, Nazli Turken, Christina Black, and Rick Smith.
10/10 we'd do it again!