Sorelle Friedler
From Wikipedia, the free encyclopedia
Alphabet Inc
Sorelle Alaina Friedler | |
|---|---|
 Friedler in 2015 | |
| Alma mater | Swarthmore College University of Maryland, College Park |
| Scientific career | |
| Institutions | Haverford College Alphabet Inc |
| Thesis | Geometric algorithms for objects in motion (2011) |
Sorelle Alaina Friedler is an American computer scientist who is an associate professor at Haverford College. She is the co-founder Association for Computing Machinery Conference on Fairness, Accountability, and Transparency. Her research seeks to prevent discrimination in machine learning.
Friedler earned her bachelor's degree at Swarthmore College.[1] She moved to the University of Maryland, College Park for her graduate studies, where she studied geometric algorithms.[2]
Research and career
Friedler joined Alphabet Inc. as a software engineer,[1][3] where she worked with X on the development of weather balloons that can provide internet access to remote communities.[1]
Friedler has advocated for the careful use of artificial intelligence and machine learning.[4] In particular, she has spoken about how biased data and algorithms reinforce social inequality.[4] In 2015 she was made a Fellow at the Data & Society Research Institute.[citation needed]
Friedler has worked with Josh Schrier and Alexander Norquist on the application of data mining to accelerate materials discovery.[5][6] They created a computer algorithm capable of predicting whether a set of reagents will create a crystalline materials when mixed in a solvent and heated.[7] To create the tool, they compiled a database of almost 4,000 chemical reactions, wrote an algorithm that could mine for patterns in data and provide insight about why some experiments fail while others succeed.[8] The algorithm was correct 89% of the time, whilst researcher (human) predictions only had a 78% success rate.[8] Friedler and her co-workers published the database online (darkreactions.haverford.edu/) to encourage other researchers to share their data.[8]
Awards and honors
Selected publications
- Feldman, Michael; Friedler, Sorelle A.; Moeller, John; Scheidegger, Carlos; Venkatasubramanian, Suresh (2015). "Certifying and Removing Disparate Impact". Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, New York, US: ACM Press. pp. 259–268. arXiv:1412.3756. doi:10.1145/2783258.2783311. ISBN 978-1-4503-3664-2. S2CID 2077168.
- "Machine Learning-Assisted Discovery of Solid Li-Ion Conducting Materials". doi:10.1021/acs.chemmater.8b03272.s001.
{{cite journal}}: Cite journal requires|journal=(help) - Adler, Philip; Falk, Casey; Friedler, Sorelle A.; Rybeck, Gabriel; Scheidegger, Carlos; Smith, Brandon; Venkatasubramanian, Suresh (2016). "Auditing Black-Box Models for Indirect Influence". 2016 IEEE 16th International Conference on Data Mining (ICDM). IEEE. pp. 1–10. arXiv:1602.07043. doi:10.1109/icdm.2016.0011. ISBN 978-1-5090-5473-2.
