I am a research scientist at the Machine Learning Research Group, Oracle Labs.

Previously, I received a Ph.D. in Computer Science from Vanderbilt University, where I was advised by Yevgeniy Vorobeychik. I graduated with B.Tech. in Electrical Engineering from Indian Institute of Technology, Kharagpur. I interned at Apple Inc. during summer 2017 and at Max Planck Institute during summer 2011. My research interests span fairness-aware machine learning and natural language processing, stochastic planning and computational game theory.

Publications

• Don’t Just Clean It, Proxy Clean It: Mitigating Bias by Proxy in Pre-Trained Models.
  Swetasudha Panda, Ari Kobren, Michael Wick, Qinlan Shen.
  Findings of Empirical Methods in Natural Language Processing (EMNLP 2022). [PDF][BibTex]

  
  

• Upstream Mitigation Is Not All You Need: Testing the Bias Transfer Hypothesis in Pre-Trained Language Models.
  Ryan Steed, Swetasudha Panda, Ari Kobren, Michael Wick.
  Association for Computational Linguistics (ACL 2022). [PDF][BibTex]

  
  

• Debiasing Pre-Trained Sentence Encoders with Word Dropout.
  Swetasudha Panda, Michael Wick, Ari Kobren.
  NeurIPS Data-Centric AI Workshop (DCAI 2021).

  
  

• Online Post-Processing In Rankings For Fair Utility Maximization.
  Ananya Gupta, Eric Johnson, Justin Payan, Aditya Roy, Ari Kobren, Swetasudha Panda, Michael Wick, Jean-Baptiste Tristan.
  Web Search And Data Mining (WSDM 2021). [PDF][BibTex]

  
  

• Fair Online Post-Processing For Black-Box ML Screening Systems.
  Swetasudha Panda, Ari Kobren, Michael Wick, Jean-Baptiste Tristan.
  Women In Machine Learning Workshop (WiML 2020). [Poster]

  
  

• Unlocking Fairness: A Trade-off Revisited.
  Michael L. Wick, Swetasudha Panda,Jean-Baptiste Tristan.
  Neural Information Processing Systems (NeurIPS 2019). [PDF] [BibTex]

  
  

• Using Bayes Factors To Control For Fairness: Case Study On Learning To Rank.
  Swetasudha Panda, Jean-Baptiste Tristan, Michael Wick, Haniyeh Mahmoudian and Pallika Kanani.
  NeurIPS 2019 Workshop on Robust AI in Financial Services: Data, Fairness, Explainability, Trustworthiness, and Privacy (Robust AI in FS 2019). [PDF]

  
  

• Scaling Hierarchical Coreference With Homomorphic Compression.
  Michael L. Wick, Swetasudha Panda, Joseph Tassarotti, Jean-Baptiste Tristan.
  Automated Knowledge Base Construction (AKBC 2019). [PDF][BibTex]

  
  

• Game Theoretic Antibody Design.
  Swetasudha Panda, Alexander M. Sevy, James E. Crowe Jr, Jens Meiler and Yevgeniy Vorobeychik.
  Joint Workshop on Autonomous Agents for Social Good (AASG 2019). [PDF] [Talk]

  
  

• Scalable Initial State Interdiction For Factored MDPs.
  Swetasudha Panda.
  Women In Machine Learning Workshop (WiML 2018).

  
  

• Algorithms For Large Scale Adversarial Decision Problems.
  Swetasudha Panda. Ph.D. Dissertation.

  
  

• Scalable Initial State Interdiction For Factored MDPs.
  Swetasudha Panda and Yevgeniy Vorobeychik.
  International Joint Conference on Artificial Intelligence (IJCAI 2018). [PDF][BibTex]

  
  

• Integrating Machine Learning With Structural Modeling To Increase HIV Neutralization Breadth.
  Swetasudha Panda, Alexander M. Sevy, James E. Crowe, Jr, Jens Meiler, Yevgeniy Vorobeychik.
  PLOS Computational Biology Journal (PLOS CompBio 2018). [PDF][BibTex]

  
  

• Near-optimal Interdiction Of Factored MDPs.
  Swetasudha Panda and Yevgeniy Vorobeychik.
  Uncertainty in Artificial Intelligence (UAI 2017). [PDF][BibTex]

  
  

• Stackelberg Games For Vaccine Design.
  Swetasudha Panda and Yevgeniy Vorobeychik.
  Autonomous Agents and MultiAgent Systems (AAMAS 2015). [PDF][BibTex]

  
  

• Designing Vaccines That Are Robust To Virus Escape.
  Swetasudha Panda and Yevgeniy Vorobeychik.
  Conference on Artificial Intelligence (AAAI 2015) (Extended Abstract) and in AAAI 2015 Spring Symposium on Applied Computational Game Theory. [PDF][BibTex]

  
  

• Robust Optic Nerve Segmentation On Clinically Acquired CT.
  Swetasudha Panda, Andrew J. Asman, Michael P. DeLisi, Louise A. Mawn, Robert L. Galloway, Bennett A. Landman.
  Conference on International Society for Optics and Photonics (SPIE 2014). [PDF][BibTex]

  
  

• Evaluation Of Multi-atlas Label Fusion For MRI Orbital Segmentation.
  Swetasudha Panda, Andrew J. Asman, Shweta P. Khare, Lindsey Thompson, Louise A. Mawn, Seth A. Smith, Bennett A. Landman.
  Journal of Medical Imaging (JMI 2014). [PDF][BibTex]

  
  

• Robust Optic Nerve Segmentation On Clinically Acquired CT.
  Swetasudha Panda, Robert A. Harrigan, Andrew J. Asman, Michael P. DeLisi, Benjamin C. W. Yvernault, Robert L. Galloway, Louise A. Mawn, Bennett A. Landman.
  Journal of Medical Imaging (JMI 2014). [PDF][BibTex]

  
  

• Regression Forest Region Recognition Enhances Multi-atlas Spleen Labeling.
  Bo Li, Swetasudha Panda, Zhoubing Xu, Andrew J. Asman, Peter L. Shanahan, Richard G. Abramson, Bennett A. Landman.
  MICCAI Challenge Workshop on Segmentation: Algorithms, Theory and Applications (MICCAI SATA 2013). [PDF][BibTex]

  
  

Program Committee/ Service

• Conference on Artificial Intelligence. AAAI'22/21/20/14
• Socially Responsible Machine Learning Workshop. ICLR'22
• BlackboxNLP Workshop. EMNLP'22/21
• ACM Conference on Fairness, Accountability and Transparency. FAccT'22/23
• AAAI/ACM Conference on Artificial Intelligence, Ethics and Society. AIES'22
• Autonomous Agents and MultiAgent Systems. AAMAS'21/17
• Women in Machine Learning Workshop. WiML'21/19
• Responsible AI Workshop. ICLR'21
• International Joint Conference on Artificial Intelligence. IJCAI'18
• Economics and Computation. EC'18/16
• ACM Workshop on Artificial Intelligence and Security. AISec'17
• Uncertainty in Artificial Intelligence. UAI'14