Measuring What Matters: Why Senior Data Scientists Look Past Accuracy
We’re continuing our new workshop series in collaboration between GirlsWhoML and Mentor Me Collective, bringing together practitioners to make AI more accessible, practical, and career-relevant.
In this live training, Neelima Verma will explore why senior data scientists look beyond accuracy when evaluating machine learning models.
Using her FairLens project as a real-world case study, Neelima will show how standard fairness checks can miss real harm, why model errors need to be measured by impact, and how to think about fairness in practical, career-relevant terms.
This session is especially useful for students, early-career data scientists, ML practitioners, and anyone interested in responsible AI, model evaluation, fairness-aware ML, or model risk.
You’ll learn:
Why accuracy is not enough in real-world ML systems
How fairness metrics can hide unequal outcomes.
How to think about model errors in terms of real-world cost.
What senior data scientists ask before deciding whether a model is ready to ship.