The Responsible Data Science Lab at Purdue investigates the tough challenges that arise when data, algorithms and real-world decisions meet. From large-scale data integrations to complex machine-learning systems, the lab works to make data-driven outcomes explainable, fair, and accountable.

We are particularly interested in:

• Uncovering hidden sources of bias and mistakes in ML pipelines — for example, tracing unexpected or discriminatory behavior back to particular components of a system and helping build tools to remediate them.
• Improving data quality and integration workflows — including how feedback from users and outcomes can be used to shift how data is cleaned, merged and prepared for analytics.
• Designing systems that connect the often-isolated stages of a data science workflow (data prep → model building → deployment/monitoring) and aligning all of them around reliable outcomes, not just isolated optimizations.