Publications

Papers here are roughly organized as:

** algorithms/optimization/theory

** applications (vision, robotics, biology,...)

** workshop papers, technical reports, etc

Preprints and Working Papers

• Y. Arjevani, A. Daniely, S. Jegelka, H. Lin. On the Complexity of Minimizing Convex Finite Sums Without Using the Indices of the Individual Functions.

Press

• Can computers help us synthesize new materials? (MIT News)
• Fabrication of new materials: Designing "recipes" using artificial intelligence (Energy Futures)
• Data diversity (MIT News | Comm. ACM)

Machine Learning algorithms / optimization / theory

• C.-Y. Chuang, J. Robinson, L. Yen-Chen, A. Torralba, S. Jegelka.Debiased Contrastive Learning. Neural Information Processing Systems (NeurIPS), 2020. Spotlight (arXiv)
• K. Gatmiry, M. Aliakbarpour, S. Jegelka. Testing Determinantal Point Processes. Neural Information Processing Systems (NeurIPS), 2020. Spotlight (arXiv)
• Y. Arjevani, J. Bruna, B. Can, M. Gürbüzbalaban, S. Jegelka, H. Lin. IDEAL: Inexact DEcentralized Accelerated Augmented Lagrangian Method. Neural Information Processing Systems (NeurIPS), 2020. Spotlight (arXiv)
• S. Curi, K.Y. Levy, S. Jegelka, A. Krause. Adaptive Sampling for Stochastic Risk-Averse Learning. Neural Information Processing Systems (NeurIPS), 2020.
• J. Robinson, S. Jegelka, S. Sra. Strength from Weakness: Fast Learning Using Weak Supervision. International Conference on Machine Learning (ICML), 2020.
• V. K. Garg, S. Jegelka, T. Jaakkola. Generalization and Representational Limits of Graph Neural Networks.. International Conference on Machine Learning (ICML), 2020.
• C.-Y. Chuang, A. Torralba, S. Jegelka. Estimating Generalization under Distribution Shifts via Domain-Invariant Representations. International Conference on Machine Learning (ICML), 2020. (project)
• M. El Halabi, S. Jegelka. Minimizing approximately submodular functions. International Conference on Machine Learning (ICML), 2020.
• J. Zhang, H. Lin, S. Jegelka, A. Jadbabaie, S. Sra. On Complexity of Finding Stationary Points of Nonsmooth Nonconvex Functions. International Conference on Machine Learning (ICML), 2020.
• J. Kirschner, I. Bogunovic, S. Jegelka, A. Krause. Distributionally Robust Bayesian Optimization. International Conference on Artificial Intelligence and Statistics (AISTATS), 2020. (arXiv)
• K. Xu, J. Li, M. Zhang, S. Du, K. Kawarabayashi, S. Jegelka. What Can Neural Networks Reason About? International Conference on Learning Representations (ICLR), 2020. Spotlight (arXiv)
• J. Robinson, S. Sra, S. Jegelka. Flexible Modeling of Diversity with Strongly Log-Concave Distributions. Neural Information Processing Systems (NeurIPS), 2019. (arXiv)
• M. Staib, S. Jegelka. Distributionally Robust Optimization and Generalization in Kernel Methods. Neural Information Processing Systems (NeurIPS), 2019. (arXiv)
• C. Bunne, D. Alvarez Melis, A. Krause, S. Jegelka. Learning Generative Models across Incomparable Spaces. International Conference on Machine Learning (ICML), 2019. (Code)
• M. Staib, S. Jegelka. Robust Budget Allocation via Continuous Submodular Functions. Applied Mathematics and Optimization, Special issue on Optimization for Data Sciences. Accepted, 2019.
• D. Alvarez-Melis, S. Jegelka and T. Jaakkola. Towards Optimal Transport with Global Invariances. International Conference on Artificial Intelligence and Statistics (AISTATS), 2019.
• M. Staib, B. Wilder and S. Jegelka. Distributionally Robust Submodular Maximization. International Conference on Artificial Intelligence and Statistics (AISTATS), 2019.
• K. Xu, W. Hu, J. Leskovec and S. Jegelka. How Powerful are Graph Neural Networks? International Conference on Learning Representations (ICLR), 2019. Oral Presentation
• G. Shulkind, S. Jegelka and G.W. Wornell. Sensor Array Design Through Submodular Optimization. IEEE Transactions on Information Theory (accepted), 2018.
• H. Lin, S. Jegelka. ResNet with one-neuron hidden layers is a Universal Approximator. Neural Information Processing Systems (NIPS), 2018. Spotlight
• I. Bogunovic, J. Scarlett, S. Jegelka, V. Cevher. Adversarially Robust Optimization with Gaussian Processes. Neural Information Processing Systems (NIPS), 2018. Spotlight
• Z. Mariet, S. Sra, S. Jegelka. Exponentiated Strongly Rayleigh Distributions. Neural Information Processing Systems (NIPS), 2018.
• J. Djolonga, S. Jegelka, A. Krause. Provable Variational Inference for Constrained Log-Submodular Models. Neural Information Processing Systems (NIPS), 2018.
• A. Gkotovos, H. Hassani, A. Krause, S. Jegelka. Discrete Sampling using Semigradient-based Product Mixtures. Conference on Uncertainty in Artificial Intelligence (UAI), 2018. Oral presentation
• K. Xu, C. Li, Y. Tian, T. Sonobe, K. Kawarabayashi, S. Jegelka. Representation Learning on Graphs with Jumping Knowledge Networks. International Conference on Machine Learning (ICML), 2018. Long talk.
• D. Alvarez-Melis, T.S. Jaakkola and S. Jegelka. Structured Optimal Transport. International Conference on Artificial Intelligence and Statistics (AISTATS), 2018. Oral presentation
• Z. Wang, C. Gehring, P. Kohli, S. Jegelka. Batched Large-scale Bayesian Optimization in High-dimensional Spaces (arXiv title: Ensemble Bayesian Optimization). International Conference on Artificial Intelligence and Statistics (AISTATS), 2018.
• B. Mirzasoleiman, S. Jegelka, A. Krause. Streaming Non-monotone Submodular Maximization: Personalized Video Summarization on the Fly. AAAI Conference on Artificial Intelligence (AAAI), 2018.
• M. Staib, S. Claici, J. Solomon, S. Jegelka. Parallel Streaming Wasserstein Barycenters. Neural Information Processing Systems (NIPS), 2017.
• C. Li, S. Jegelka, S. Sra. Column Subset Selection via Polynomial Time Dual Volume Sampling. Neural Information Processing Systems (NIPS), 2017.
• M. Staib, S. Jegelka. Robust Budget Allocation via Continuous Submodular Functions. International Conference on Machine Learning (ICML), 2017.
• Z. Wang, S. Jegelka. Max-value Entropy Search for Efficient Bayesian Optimization. International Conference on Machine Learning (ICML), 2017. (Code)
• Z. Wang, C. Li, S. Jegelka, P. Kohli. Batched High-dimensional Bayesian Optimization via Structural Kernel Learning. International Conference on Machine Learning (ICML), 2017. (Code)
• C. Li, S. Sra, S. Jegelka. Fast Mixing Markov Chains for Strongly Rayleigh Measures, DPPs, and Constrained Sampling. Neural Information Processing Systems (NIPS), 2016. (Part I on arXiv: Fast Sampling for Strongly Rayleigh Measures with Application to Determinantal Point Processes)
• J. Djolonga, S. Jegelka, S. Tschiatschek, A. Krause. Cooperative Graphical Models. Neural Information Processing Systems (NIPS), 2016. (video spotlight)
• S. Jegelka and J. Bilmes. Graph Cuts with Interacting Edge Costs - Examples, Approximations, and Algorithms. Mathematical Programming Ser. A, 162:241-282, 2017. (arXiv version)
• C. Li, S. Sra, S. Jegelka. Gaussian quadrature for matrix inverse forms with applications. International Conference on Machine Learning (ICML), 2016. (code)
• C. Li, S. Jegelka, S. Sra. Fast DPP Sampling for Nyström with Application to Kernel Methods. International Conference on Machine Learning (ICML), 2016. (code)
• C. Li, S. Jegelka, S. Sra. Efficient Sampling for k-Determinantal Point Processes. Conference on Artificial Intelligence and Statistics (AISTATS) 2016, oral presentation.
• Z. Wang, B. Zhou, S. Jegelka. Optimization as Estimation with Gaussian Processes in Bandit Settings. Conference on Artificial Intelligence and Statistics (AISTATS) 2016, oral presentation. (code)
• X. Pan, S. Jegelka, J. Gonzalez, J. Bradley and M.I. Jordan. Parallel Double Greedy Submodular Maximization. Neural Information Processing Systems (NIPS), 2014. (supplement | code | project page)
• R. Nishihara, S. Jegelka and M.I. Jordan. On the Linear Convergence Rate of Decomposable Submodular Function Minimization. Neural Information Processing Systems (NIPS), 2014. (talk)
• A. Prasad, S. Jegelka and D. Batra. Submodular meets Structured: Finding Diverse Subsets in Exponentially-Large Structured Item Sets. Neural Information Processing Systems (NIPS), 2014. Spotlight
• R. Iyer, S. Jegelka and J. Bilmes. Monotone Closure of Relaxed Constraints in Submodular Optimization: Connections Between Minimization and Maximization. Conference on Uncertainty in Artificial Intelligence (UAI), 2014.
• X. Pan, J. Gonzalez, S. Jegelka, T. Broderick and M.I. Jordan. Optimistic Concurrency Control for Distributed Unsupervised Learning. Neural Information Processing Systems (NIPS), 2013. (extended version | project page)
• S. Jegelka, F. Bach and S. Sra. Reflection methods for user-friendly submodular optimization. Neural Information Processing Systems (NIPS), 2013. (talk | extended version | some code)
• R. Iyer, S. Jegelka and J. Bilmes. Curvature and Optimal Algorithms for Learning and Minimizing Submodular Functions. Neural Information Processing Systems (NIPS), 2013. (extended version)
• R. Iyer, S. Jegelka and J. Bilmes. Fast Semidifferential-based Submodular Function Optimization. International Conference on Machine Learning (ICML), 2013. (supplement | code) Best paper award
• S. Jegelka, H. Lin and J. Bilmes. On Fast Approximate Submodular Minimization. 25th Annual Conference on Neural Information Processing Systems (NIPS), 2011.
• S. Jegelka and J. Bilmes. Online Submodular Minimization for Combinatorial Structures. 28th International Conference on Machine Learning (ICML), 2011. (Supplementary material)
• S. Jegelka and J. Bilmes. Approximation Bounds for Inference using Cooperative Cut. 28th International Conference on Machine Learning (ICML), 2011.
• H. Shen, S. Jegelka and A. Gretton. Fast Kernel-based Independent Component Analysis. IEEE Transactions on Signal Processing 57(9), pp. 3498-3511, 2009.
• S. Jegelka, S. Sra and A. Banerjee. Approximation Algorithms for Tensor Clustering. Algorithmic Learning Theory: 20th International Conference (ALT), 2009.
• S. Nowozin and S. Jegelka. Solution Stability in Linear Programming Relaxations: Graph Partitioning and Unsupervised Learning. 26th International Conference on Machine Learning (ICML), 2009.
• S. Jegelka, A. Gretton, B. Schoelkopf, B.K. Sriperumbudur and U. von Luxburg. Generalized Clustering via Kernel Embeddings. KI 2009: Advances in Artificial Intelligence, 2009.
• U. von Luxburg, S. Bubeck, S. Jegelka and M. Kaufmann. Consistent Minimization of Clustering Objective Functions. 21st Annual Conference on Neural Information Processing Systems (NIPS), 2007.
• H. Shen, S. Jegelka and A. Gretton. Fast Kernel ICA using an Approximate Newton Method. 11th Conference on Artificial Intelligence and Statistics (AISTATS), 2007.
• S. Jegelka and A. Gretton. Brisk Kernel Independent Component Analysis. In L. Bottou, O. Chapelle, D. DeCoste, J. Weston, editors. Large Scale Kernel Machines, pp. 225-250. MIT Press, 2007.

Applications

(computer vision, computational biology, computational neuroscience, and more)

• E. Kim, Z. Jensen, A. van Grootel, K. Huang, M. Staib, S. Mysore, H.-S. Chang, E. Strubell, A. McCallum, S. Jegelka, E. Olivetti. Inorganic Materials Synthesis Planning with Literature-Trained Neural Networks. Journal of Chemical Information and Modeling, 2020. (arXiv)
• M. Zhang, K. Xu, K. Kawarabayashi, S. Jegelka and J. Boyd-Graber. Are Girls Neko or Shojo? Cross-Lingual Mapping of Non-Isomorphic Embedding with Iterative Normalization. ACL Short paper, 2019. (arXiv)
• E. Kim, K. Huang, S. Jegelka, E. Olivetti. Virtual screening of inorganic materials synthesis parameters with deep learning. npj Computational Materials, accepted
• H. Song, S. Jegelka, V. Rathod and K. Murphy. Deep Metric Learning via Facility Location. International Conference on Computer Vision and Pattern Recognition (CVPR), 2017. Spotlight
• Z. Wang, S. Jegelka, L. P. Kaelbling, T. Lozano-Perez. Focused Model-Learning and Planning for Non-Gaussian Continuous State-Action Systems. IEEE International Conference on Robotics and Automation (ICRA), 2017.
• G. Shulkind, S. Jegelka and G. W. Wornell. Multiple wavelength sensing array design. ICASSP 2017.
• H. Song, Y. Xiang, S. Jegelka and S. Savarese. Deep Metric Learning via Lifted Structured Feature Embedding. International Conference on Computer Vision and Pattern Recognition (CVPR), 2016. Spotlight
• S. Azadi, J. Feng, S. Jegelka, T. Darrell. Auxiliary Image Regularization for Deep CNNs with Noisy Labels. International Conference on Learning Representations (ICLR) 2016.
• A. Prasad, S. Jegelka and D. Batra. Submodular meets Structured: Finding Diverse Subsets in Exponentially-Large Structured Item Sets. Neural Information Processing Systems (NIPS), 2014. Spotlight
• H. Song, Y.J. Lee, S. Jegelka and T. Darrell. Weakly-supervised Discovery of Visual Pattern Configurations. Neural Information Processing Systems (NIPS), 2014.
• V. Strnadova, A. Buluc, J. Chapman, J. Gonzalez, S. Jegelka, J. Gilbert, D. Rokhsar and L. Oliker. Efficient and Accurate Clustering for Large-Scale Genetic Mapping. IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2014.
• H. Song, R. Girshick, S. Jegelka, J. Mairal, Z. Harchaoui and T. Darrell. On learning to localize objects with minimal supervision. International Conference on Machine Learning (ICML), 2014. (talk | code)
• J. Feng, S. Jegelka, S. Yang and T. Darrell. Learning Scalable Discriminative Dictionaries with Sample Relatedness. IEEE Conference of Computer Vision and Pattern Recognition (CVPR), 2014 oral presentation. (supplement)
• S. Jegelka, A. Kapoor and E. Horvitz. An interactive approach to solving correspondence problems. International Journal of Computer Vision, 2013.
• P. Kohli, A. Osokin and S. Jegelka. A principled deep random field model for image segmentation. IEEE Conference of Computer Vision and Pattern Recognition (CVPR), 2013 (supplement | code)
• S. Jegelka and J. Bilmes. Submodularity beyond Submodular Energies: Coupling Edges in Graph Cuts. IEEE Conference of Computer Vision and Pattern Recognition (CVPR), 2011 oral presentation, (3.5% acceptance rate). (supplementary material | new code (by Evan Shelhamer) | old code | data)
• S. Jegelka, J. A. Bednar and R. Miikkulainen. Prenatal Development of Ocular Dominance in a Self-organizing Model of V1. Neurocomputing 69, pp. 1291-1296, 2006.

Abstracts and Contributions to Workshops

• Z. Xu. C. Li, S. Jegelka. Exploring the Robustness of GANs to Internal Perturbations. ICML workshop on Security and Privacy on Machine Learning, 2019. (arXiv)
• C.-Y. Chuang, A. Torralba, S. Jegelka. The Role of Embedding-complexity in Domain-invariant Representations. ICML workshop on Adaptive and Multitask Learning, 2019.
• C. Bunne, D. Alvarez Melis, A. Krause and S. Jegelka. Learning Generative Models Across Incomparable Spaces. NIPS workshop on Relational Representation Learning, 2018. Best Paper Award, Oral presentation
• M. Staib, B. Wilder and S. Jegelka. Distributionally Robust Submodular Maximization. ICML 2018 Workshop on Modern Trends in Nonconvex Optimization for Machine Learning.Spotlight.
• M. Staib and S. Jegelka. Distributionally Robust Deep Learning as a Generalization of Adversarial Training. NIPS Machine Learning and Computer Security Workshop, 2017.
• A. Lenail, L. Schmidt, J. Li, T. Ehrenberger, K. Sachs, S. Jegelka and E. Fraenkel. Graph-Sparse Logistic Regression. NIPS workshop on Discrete Structure in Machine Learning (DISCML), 2017.
• Z. Wang, C. Gehring, P. Kohli, S. Jegelka. Batched Large-scale Bayesian Optimization in High-dimensional Spaces. NIPS workshop on Bayesian Optimization (BayesOpt), 2017.
• D. Alvarez Melis, T. Jaakkola and S. Jegelka. Structured Optimal Transport. NIPS workshop on Optimal Transport and Machine Learning (OTML), 2017. Oral presentation
• M. Cohen, L. Schmidt, C. Hegde, S. Jegelka. Efficiently Optimizing over (Non-Convex) Cones via Approximate Projections. NIPS workshop on Optimization in Machine Learning (OPTML), 2017. Oral presentation.
• M. Staib and S. Jegelka. a href="http://www.mit.edu/~mstaib/ci2017_with_proof.pdf">Wasserstein k-means++ for Cloud Regime Histogram Clustering. Climate Informatics, 2017.
• Z. Wang, B. Zhou and S. Jegelka. Optimization as Estimation with Gaussian Processes in Bandit Settings. NIPS workshop on Bayesian Optimization: Scalability and Flexibility, 2015.
• E. Shelhamer, S. Jegelka and T. Darrell. Communal Cuts: sharing cuts across images. NIPS workshop on Discrete Optimization in Machine Learning, 2014.
• V. Strnadova, A. Buluc, L. Oliker, J. Gonzalez, S. Jegelka, J. Chapman and J. R. Gilbert. Fast Clustering Methods for Genetic Mapping in Plants. 16th SIAM Conference on Parallel Processing for Scientific Computing, 2014.
• R. Iyer, S. Jegelka and J. Bilmes. Mirror Descent-Like Algorithms for Submodular Optimization. NIPS 2012 Workshop on Discrete Optimization in Machine Learning.
• S. Jegelka and J. Bilmes. Multi-label Cooperative Cuts. CVPR 2011 Workshop on Inference in Graphical Models with Structured Potentials.
• S. Jegelka and J. Bilmes. Coupling Edges in Graph Cuts. SIAM Conference on Optimization, 2011.
• S. Jegelka and J. Bilmes. Online Algorithms for Submodular Minimization with Combinatorial Constraints. NIPS 2010 Workshop Discrete Optimization in Machine Learning.
• S. Jegelka and J. Bilmes. Cooperative Cuts: Graph Cuts with Submodular Edge Weights. EURO XXIV (24th European Conference on Operational Research), 2010.
• S. Jegelka and J. Bilmes. Notes on Graph Cuts with Submodular Edge Weights. NIPS 2009 Workshop Discrete Optimization in Machine Learning.
• S. Jegelka, A. Gretton and D. Achlioptas. Kernel ICA for Large Scale Problems. NIPS 2005 Workshop on Large Scale Kernel Machines.

Technical Reports

• C. Li, D. Alvarez-Melis, K. Xu, S. Jegelka and S. Sra. Distributional Adversarial Networks (Code)
• A. Kapoor, E.P. Frady, S. Jegelka, W.B. Kristan, and E. Horvitz. Inferring and Learning from Neuronal Correspondences. 2015.
• V. Strnadova, A. Buluc, J. Chapman, J. R. Gilbert, J. Gonzalez, S. Jegelka, D. Rokhsar and L. Oliker. Efficient and Accurate Clustering for Large-Scale Genetic Mapping. UC Santa Barbara, Dept. of CS, TR UCSB-2014-03.
• S. Jegelka and J. Bilmes. Cooperative Cuts for Image Segmentation, UWEETR-1020-0003, University of Washington, 2010.
• S. Jegelka and J. Bilmes. Cooperative Cuts: Graph Cuts with Submodular Edge Weights. MPI-TR 189, 2010.
• S. Sra, S. Jegelka and A. Banerjee. Approximation Algorithms for Bregman Clustering, Co-Clustering and Tensor Clustering. MPI-TR 177, 2008.
• B. Kulis, S. Sra and S. Jegelka. Scalable Semidefinite Programming using Convex Perturbations. TR 07-47, University of Texas at Austin, 2007.
• H. Shen, S. Jegelka and A. Gretton. Geometric Analysis of Hilbert Schmidt Independence Criterion based ICA contrast function. TR PA006080, NICTA, 2006.

Theses

• Combinatorial Problems with Submodular Coupling in Machine Learning and Computer Vision. Ph.D. Thesis, ETH Zurich, 2012.
• Statistical Learning Theory Approaches to Clustering. Diplom Thesis, Universitaet Tuebingen, 2007.
• Stochastic Approximations for Large-Scale Kernel ICA. Studienarbeit, Universitaet Tuebingen, 2006.