PNNL

Sense of inclusion, empowerment, and networking are some of the top reasons Emilie Purvine, a senior data scientist at Pacific Northwest National Laboratory (PNNL), eagerly participates in research-collaboration workshops for women.

Purvine discovered multiple workshops and networks through the Association for Women in Mathematics website. Within these spaces, women in mathematics, computer science, and data science come together from different stages of their careers to collaborate on projects. They focus on cutting-edge problems in areas such as predictive modeling, statistical and topological learning, and multiscale representation. These areas align with PNNL’s laboratory objectives and more specifically with the Artificial Intelligence and Data Analytics Division that Purvine is a part of.

Purvine shared that one of her favorite experiences was from her participation in a 2016 Women in Computational Topology (WinCompTop) event. It is within such teams that bonds are made, as she gained a sense of inclusion.

“Being grouped with those who self-identify as women creates a different type of atmosphere. I feel there is more understanding, patience, encouragement, and sense of being heard,” said Purvine.

One of the best parts is that connections don’t have to stop at the end of the event. Participants are encouraged to exchange contact information so they can continue to collaborate. For Purvine’s 2016 WinCompTop team of seven, they are spread across the United States but that has not stopped them from holding regular meetings in person and virtually over the years.

“While yes, we continue to make progress on our project and have published five papers over the years, we also continue to empower each other. Within my team, we have grown in our careers and in our personal lives. We have become wives and mothers, all while remaining passionate and encouraging toward our careers,” said Purvine.

Seeking to continue networking and working on projects of interest, Purvine signed up for the 2019 Women in the Science of Data and Mathematics (WiSDM) event and extended the invitation to her fellow PNNL coworkers, including data scientist Brenda Praggastis. Purvine’s expanded WiSDM team crossed national boundaries this time and includes a member from Turkey. She shared that they continue to advance their project, “Learning Temporal Representations of Dynamic Networks for Anomaly Detection and Community Discovery,” exploring methods at the intersection of machine learning and linear algebra.

Purvine explained that thanks to small collaboration grants from math institutes, it’s possible for their project to advance over the years. The team has researched and applied to different funding opportunities they found through the Mathematical Sciences Institutes so that members could afford to meet in person. By communicating the importance of the research and in-person collaboration, her team has met two additional times after the original WiSDM event.

“There is so much more progress that happens during back-to-back, in-person sessions that we just can’t achieve in multiple, staggered Zoom meetings. While maybe it’s not one major leap, the week consists of steady progress, so it feels like a significant step forward at the end,” said Purvine.

As she becomes more experienced in her field, Purvine shared that she has likely “aged-out” of mentee opportunities at some of the workshops. However, she continues to keep her network active because she would love to have the chance to come full circle and pay it forward as a mentor.