Bio


I am an Assistant Professor in the Department of Computer Science at Stanford University, where I am affiliated with the Artificial Intelligence Laboratory and a fellow of the Woods Institute for the Environment.

My research is centered on techniques for scalable and accurate inference in graphical models, statistical modeling of data, large-scale combinatorial optimization, and robust decision making under uncertainty, and is motivated by a range of applications, in particular ones in the emerging field of computational sustainability.

Academic Appointments


Honors & Awards


  • Sloan Research Fellowship, Alfred P. Sloan Foundation
  • IJCAI Computers and Thought Award, IJCAI
  • Microsoft Research Faculty Fellowship, Microsoft Research
  • NSF CAREER Award, National Science Foundation
  • ONR Young Investigator Award, Office of Naval Research
  • AFOSR Young Investigator Award, Air Force Office of Scientific Research
  • AWS Machine Learning Research Award, Amazon Web Services (AWS)
  • Sony Faculty Innovation Award, Sony
  • Hellman Fellowship, Hellman Foundation
  • AAAI 2017 Outstanding Paper Award, AAAI
  • Bloomberg Data Science Research Grant, Bloomberg
  • 10 World Changing Ideas of 2016, Scientific American
  • First Place, World Bank Big Data Innovation Challenge, World Bank
  • Finalist, NVIDIA Global Impact Award, NVIDIA

Professional Education


  • Ph.D., Cornell University, Computer Science (2015)

2021-22 Courses


Stanford Advisees


All Publications


  • Scalable deep learning to identify brick kilns and aid regulatory capacity. Proceedings of the National Academy of Sciences of the United States of America Lee, J., Brooks, N. R., Tajwar, F., Burke, M., Ermon, S., Lobell, D. B., Biswas, D., Luby, S. P. 2021; 118 (17)

    Abstract

    Improving compliance with environmental regulations is critical for promoting clean environments and healthy populations. In South Asia, brick manufacturing is a major source of pollution but is dominated by small-scale, informal producers who are difficult to monitor and regulate-a common challenge in low-income settings. We demonstrate a low-cost, scalable approach for locating brick kilns in high-resolution satellite imagery from Bangladesh. Our approach identifies kilns with 94.2% accuracy and 88.7% precision and extracts the precise GPS coordinates of every brick kiln across Bangladesh. Using these estimates, we show that at least 12% of the population of Bangladesh (>18 million people) live within 1 km of a kiln and that 77% and 9% of kilns are (illegally) within 1 km of schools and health facilities, respectively. Finally, we show how kilns contribute up to 20.4 mug/[Formula: see text] of [Formula: see text] (particulate matter of a diameter less than 2.5 mum) in Dhaka when the wind blows from an unfavorable direction. We document inaccuracies and potential bias with respect to local regulations in the government data. Our approach demonstrates how machine learning and Earth observation can be combined to better understand the extent and implications of regulatory compliance in informal industry.

    View details for DOI 10.1073/pnas.2018863118

    View details for PubMedID 33888583

  • Using satellite imagery to understand and promote sustainable development. Science (New York, N.Y.) Burke, M., Driscoll, A., Lobell, D. B., Ermon, S. 2021; 371 (6535)

    Abstract

    Accurate and comprehensive measurements of a range of sustainable development outcomes are fundamental inputs into both research and policy. We synthesize the growing literature that uses satellite imagery to understand these outcomes, with a focus on approaches that combine imagery with machine learning. We quantify the paucity of ground data on key human-related outcomes and the growing abundance and improving resolution (spatial, temporal, and spectral) of satellite imagery. We then review recent machine learning approaches to model-building in the context of scarce and noisy training data, highlighting how this noise often leads to incorrect assessment of model performance. We quantify recent model performance across multiple sustainable development domains, discuss research and policy applications, explore constraints to future progress, and highlight research directions for the field.

    View details for DOI 10.1126/science.abe8628

    View details for PubMedID 33737462

  • A Framework for Sample Efficient Interval Estimation with Control Variates Zhao, S., Yeh, C., Ermon, S., Chiappa, S., Calandra, R. ADDISON-WESLEY PUBL CO. 2020: 4583–91
  • AlignFlow: Cycle Consistent Learning from Multiple Domains via Normalizing Flows Grover, A., Chute, C., Shu, R., Cao, Z., Ermon, S., Assoc Advancement Artificial Intelligence ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2020: 4028-4035
  • Meta-Amortized Variational Inference and Learning Wu, M., Choi, K., Goodman, N., Ermon, S., Assoc Advancement Artificial Intelligence ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2020: 6404-6412
  • Using publicly available satellite imagery and deep learning to understand economic well-being in Africa. Nature communications Yeh, C. n., Perez, A. n., Driscoll, A. n., Azzari, G. n., Tang, Z. n., Lobell, D. n., Ermon, S. n., Burke, M. n. 2020; 11 (1): 2583

    Abstract

    Accurate and comprehensive measurements of economic well-being are fundamental inputs into both research and policy, but such measures are unavailable at a local level in many parts of the world. Here we train deep learning models to predict survey-based estimates of asset wealth across ~ 20,000 African villages from publicly-available multispectral satellite imagery. Models can explain 70% of the variation in ground-measured village wealth in countries where the model was not trained, outperforming previous benchmarks from high-resolution imagery, and comparison with independent wealth measurements from censuses suggests that errors in satellite estimates are comparable to errors in existing ground data. Satellite-based estimates can also explain up to 50% of the variation in district-aggregated changes in wealth over time, with daytime imagery particularly useful in this task. We demonstrate the utility of satellite-based estimates for research and policy, and demonstrate their scalability by creating a wealth map for Africa's most populous country.

    View details for DOI 10.1038/s41467-020-16185-w

    View details for PubMedID 32444658

  • Closed-loop optimization of fast-charging protocols for batteries with machine learning. Nature Attia, P. M., Grover, A. n., Jin, N. n., Severson, K. A., Markov, T. M., Liao, Y. H., Chen, M. H., Cheong, B. n., Perkins, N. n., Yang, Z. n., Herring, P. K., Aykol, M. n., Harris, S. J., Braatz, R. D., Ermon, S. n., Chueh, W. C. 2020; 578 (7795): 397–402

    Abstract

    Simultaneously optimizing many design parameters in time-consuming experiments causes bottlenecks in a broad range of scientific and engineering disciplines1,2. One such example is process and control optimization for lithium-ion batteries during materials selection, cell manufacturing and operation. A typical objective is to maximize battery lifetime; however, conducting even a single experiment to evaluate lifetime can take months to years3-5. Furthermore, both large parameter spaces and high sampling variability3,6,7 necessitate a large number of experiments. Hence, the key challenge is to reduce both the number and the duration of the experiments required. Here we develop and demonstrate a machine learning methodology  to efficiently optimize a parameter space specifying the current and voltage profiles of six-step, ten-minute fast-charging protocols for maximizing battery cycle life, which can alleviate range anxiety for electric-vehicle users8,9. We combine two key elements to reduce the optimization cost: an early-prediction model5, which reduces the time per experiment by predicting the final cycle life using data from the first few cycles, and a Bayesian optimization algorithm10,11, which reduces the number of experiments by balancing exploration and exploitation to efficiently probe the parameter space of charging protocols. Using this methodology, we rapidly identify high-cycle-life charging protocols among 224 candidates in 16 days (compared with over 500 days using exhaustive search without early prediction), and subsequently validate the accuracy and efficiency of our optimization approach. Our closed-loop methodology automatically incorporates feedback from past experiments to inform future decisions and can be generalized to other applications in battery design and, more broadly, other scientific domains that involve time-intensive experiments and multi-dimensional design spaces.

    View details for DOI 10.1038/s41586-020-1994-5

    View details for PubMedID 32076218

  • Efficient Object Detection in Large Images Using Deep Reinforcement Learning Uzkent, B., Yeh, C., Ermon, S., IEEE Comp Soc IEEE COMPUTER SOC. 2020: 1813–22
  • Cloud Removal in Satellite Images Using Spatiotemporal Generative Networks Sarukkai, V., Jain, A., Uzkent, B., Ermon, S., IEEE Comp Soc IEEE COMPUTER SOC. 2020: 1785–94
  • Permutation Invariant Graph Generation via Score-Based Generative Modeling Niu, C., Song, Y., Song, J., Zhao, S., Grover, A., Ermon, S., Chiappa, S., Calandra, R. ADDISON-WESLEY PUBL CO. 2020: 4474–83
  • Gaussianization Flows Meng, C., Song, Y., Song, J., Ermon, S., Chiappa, S., Calandra, R. ADDISON-WESLEY PUBL CO. 2020: 4336–44
  • Streamlining variational inference for constraint satisfaction problems JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT Grover, A., Achim, T., Ermon, S. 2019; 2019 (12)
  • Computational Sustainability: Computing for a Better World and a Sustainable Future COMMUNICATIONS OF THE ACM Gomes, C., Dietterich, T., Barrett, C., Conrad, J., Dilkina, B., Ermon, S., Fang, F., Farnsworth, A., Fern, A., Fern, X., Fink, D., Fisher, D., Flecker, A., Freund, D., Fuller, A., Gregoire, J., Hoperoft, J., Kelling, S., Kolter, Z., Powell, W., Sintov, N., Selker, J., Selman, B., Sheldon, D., Shmoys, D., Tambe, M., Wong, W., Wood, C., Wu, X., Xue, Y., Yadav, A., Yakubu, A., Zeeman, M. 2019; 62 (9): 56–65

    View details for DOI 10.1145/3339399

    View details for Web of Science ID 000483033900019

  • High-Voltage Charging-Induced Strain, Heterogeneity, and Micro-Cracks in Secondary Particles of a Nickel-Rich Layered Cathode Material ADVANCED FUNCTIONAL MATERIALS Mao, Y., Wang, X., Xia, S., Zhang, K., Wei, C., Bak, S., Shadike, Z., Liu, X., Yang, Y., Xu, R., Pianetta, P., Ermon, S., Stavitski, E., Zhao, K., Xu, Z., Lin, F., Yang, X., Hu, E., Liu, Y. 2019; 29 (18)
  • Using machine learning to discover shape descriptors for predicting emulsion stability in a microfluidic channel SOFT MATTER Khor, J., Jean, N., Luxenberg, E. S., Ermon, S., Tang, S. Y. 2019; 15 (6): 1361–72

    View details for DOI 10.1039/c8sm02054j

    View details for Web of Science ID 000459588200020

  • InfoVAE: Balancing Learning and Inference in Variational Autoencoders Zhao, S., Song, J., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2019: 5885–92
  • Mapping Missing Population in Rural India: A Deep Learning Approach with Satellite Imagery Hu, W., Patel, J., Robert, Z., Novosad, P., Asher, S., Tang, Z., Burke, M., Lobell, D., Ermon, S., Assoc Comp Machinery ASSOC COMPUTING MACHINERY. 2019: 353–59
  • Tile2Vec: Unsupervised Representation Learning for Spatially Distributed Data Jean, N., Wang, S., Samar, A., Azzari, G., Lobell, D., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2019: 3967–74
  • Rapid identification of pathogenic bacteria using Raman spectroscopy and deep learning. Nature communications Ho, C. S., Jean, N. n., Hogan, C. A., Blackmon, L. n., Jeffrey, S. S., Holodniy, M. n., Banaei, N. n., Saleh, A. A., Ermon, S. n., Dionne, J. n. 2019; 10 (1): 4927

    Abstract

    Raman optical spectroscopy promises label-free bacterial detection, identification, and antibiotic susceptibility testing in a single step. However, achieving clinically relevant speeds and accuracies remains challenging due to weak Raman signal from bacterial cells and numerous bacterial species and phenotypes. Here we generate an extensive dataset of bacterial Raman spectra and apply deep learning approaches to accurately identify 30 common bacterial pathogens. Even on low signal-to-noise spectra, we achieve average isolate-level accuracies exceeding 82% and antibiotic treatment identification accuracies of 97.0±0.3%. We also show that this approach distinguishes between methicillin-resistant and -susceptible isolates of Staphylococcus aureus (MRSA and MSSA) with 89±0.1% accuracy. We validate our results on clinical isolates from 50 patients. Using just 10 bacterial spectra from each patient isolate, we achieve treatment identification accuracies of 99.7%. Our approach has potential for culture-free pathogen identification and antibiotic susceptibility testing, and could be readily extended for diagnostics on blood, urine, and sputum.

    View details for DOI 10.1038/s41467-019-12898-9

    View details for PubMedID 31666527

  • Training Variational Autoencoders with Buffered Stochastic Variational Inference Shu, R., Bui, H. H., Whang, J., Ermon, S., Chaudhuri, K., Sugiyama, M. MICROTOME PUBLISHING. 2019
  • Learning Controllable Fair Representations Song, J., Kalluri, P., Grover, A., Zhao, S., Ermon, S., Chaudhuri, K., Sugiyama, M. MICROTOME PUBLISHING. 2019
  • Uncertainty Autoencoders: Learning Compressed Representations via Variational Information Maximization Grover, A., Ermon, S., Chaudhuri, K., Sugiyama, M. MICROTOME PUBLISHING. 2019
  • Differentiable Antithetic Sampling for Variance Reduction in Stochastic Variational Inference Wu, M., Goodman, N., Ermon, S., Chaudhuri, K., Sugiyama, M. MICROTOME PUBLISHING. 2019
  • Predicting Economic Development using Geolocated Wikipedia Articles Sheehan, E., Meng, C., Tan, M., Uzkent, B., Jean, N., Burke, M., Lobell, D., Ermon, S., Assoc Comp Machinery ASSOC COMPUTING MACHINERY. 2019: 2698–2706
  • Bias Correction of Learned Generative Models using Likelihood-Free Importance Weighting Grover, A., Song, J., Agarwal, A., Tran, K., Kapoor, A., Horvitz, E., Ermon, S., Wallach, H., Larochelle, H., Beygelzimer, A., d'Alche-Buc, F., Fox, E., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2019
  • MintNet: Building Invertible Neural Networks with Masked Convolutions Song, Y., Meng, C., Ermon, S., Wallach, H., Larochelle, H., Beygelzimer, A., d'Alche-Buc, F., Fox, E., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2019
  • Meta-Inverse Reinforcement Learning with Probabilistic Context Variables Yu, L., Yu, T., Finn, C., Ermon, S., Wallach, H., Larochelle, H., Beygelzimer, A., d'Alche-Buc, F., Fox, E., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2019
  • Approximating the Permanent by Sampling from Adaptive Partitions Kuck, J., Dao, T., Rezatofighi, H., Sabharwal, A., Ermon, S., Wallach, H., Larochelle, H., Beygelzimer, A., d'Alche-Buc, F., Fox, E., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2019
  • Generative Modeling by Estimating Gradients of the Data Distribution Song, Y., Ermon, S., Wallach, H., Larochelle, H., Beygelzimer, A., d'Alche-Buc, F., Fox, E., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2019
  • Temporal FiLM: Capturing Long-Range Sequence Dependencies with Feature-Wise Modulation Birnbaum, S., Kuleshov, V., Enam, S., Koh, P., Ermon, S., Wallach, H., Larochelle, H., Beygelzimer, A., d'Alche-Buc, F., Fox, E., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2019
  • Using machine learning to discover shape descriptors for predicting emulsion stability in a microfluidic channel. Soft matter Khor, J. W., Jean, N., Luxenberg, E. S., Ermon, S., Tang, S. K. 2018

    Abstract

    In soft matter consisting of many deformable objects, object shapes often carry important information about local forces and their interactions with the local environment, and can be tightly coupled to the bulk properties and functions. In a concentrated emulsion, for example, the shapes of individual droplets are directly related to the local stress arising from interactions with neighboring drops, which in turn determine their stability and the resulting rheological properties. Shape descriptors used in prior work on single drops and dilute emulsions, where droplet-droplet interactions are largely negligible and the drop shapes are simple, are insufficient to fully capture the broad range of droplet shapes in a concentrated system. This paper describes the application of a machine learning method, specifically a convolutional autoencoder model, that learns to: (1) discover a low-dimensional code (8-dimensional) to describe droplet shapes within a concentrated emulsion, and (2) predict whether the drop will become unstable and undergo break-up. The input consists of images (N = 500002) of two-dimensional droplet boundaries extracted from movies of a concentrated emulsion flowing through a confined microfluidic channel as a monolayer. The model is able to faithfully reconstruct droplet shapes, as well as to achieve a classification accuracy of 91.7% in the prediction of droplet break-up, compared with 60% using conventional scalar descriptors based on droplet elongation. It is observed that 4 out of the 8 dimensions of the code are interpretable, corresponding to drop skewness, elongation, throat size, and surface curvature, respectively. Furthermore, the results show that drop elongation, throat size, and surface curvature are dominant factors in predicting droplet break-up for the flow conditions tested. The method presented is expected to facilitate follow-on work to identify the relationship between drop shapes and the interactions with other drops, and to identify potentially new modes of break-up mechanisms in a concentrated system. Finally, the method developed here should also apply to other soft materials such as foams, gels, and cells and tissues.

    View details for PubMedID 30570628

  • Learning with Weak Supervision from Physics and Data-Driven Constraints AI MAGAZINE Ren, H., Stewart, R., Song, J., Kuleshov, V., Ermon, S. 2018; 39 (1): 27–38
  • Bias and Generalization in Deep Generative Models: An Empirical Study Zhao, S., Ren, H., Yuan, A., Song, J., Goodman, N., Ermon, S., Bengio, S., Wallach, H., Larochelle, H., Grauman, K., CesaBianchi, N., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2018
  • Variational Rejection Sampling Grover, A., Gummadi, R., Lazaro-Gredilla, M., Schuurmans, D., Ermon, S., Storkey, A., PerezCruz, F. MICROTOME PUBLISHING. 2018
  • Best arm identification in multi-armed bandits with delayed feedback Grover, A., Markov, T., Attia, P., Jin, N., Perkins, N., Cheong, B., Chen, M., Yang, Z., Harris, S., Chueh, W., Ermon, S., Storkey, A., PerezCruz, F. MICROTOME PUBLISHING. 2018
  • The Information Autoencoding Family: A Lagrangian Perspective on Latent Variable Generative Models Zhao, S., Song, J., Ermon, S., Globerson, A., Silva, R. AUAI PRESS. 2018: 1031–41
  • Bayesian optimization and attribute adjustment Eismann, S., Levy, D., Shu, R., Bartzsch, S., Ermon, S., Globerson, A., Silva, R. AUAI PRESS. 2018: 1042–52
  • End-to-End Learning of Motion Representation for Video Understanding Fan, L., Huang, W., Gan, C., Ermon, S., Gong, B., Huang, J., IEEE IEEE. 2018: 6016–25
  • Semi-supervised Deep Kernel Learning: Regression with Unlabeled Data by Minimizing Predictive Variance Jean, N., Xie, S., Ermon, S., Bengio, S., Wallach, H., Larochelle, H., Grauman, K., CesaBianchi, N., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2018
  • Infrastructure Quality Assessment in Africa using Satellite Imagery and Deep Learning Oshri, B., Hu, A., Adelson, P., Chen, X., Dupas, P., Weinstein, J., Burke, M., Lobell, D., Ermon, S., ACM ASSOC COMPUTING MACHINERY. 2018: 616–25
  • Deep Transfer Learning for Crop Yield Prediction with Remote Sensing Data Wang, A. X., Tran, C., Desai, N., Lobell, D., Ermon, S., Assoc Comp Machinery ASSOC COMPUTING MACHINERY. 2018
  • Amortized Inference Regularization Shu, R., Bui, H. H., Zhao, S., Kochenderfer, M. J., Ermon, S., Bengio, S., Wallach, H., Larochelle, H., Grauman, K., CesaBianchi, N., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2018
  • Flow-GAN: Combining Maximum Likelihood and Adversarial Learning in Generative Models Grover, A., Dhar, M., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2018: 3069–76
  • Boosted Generative Models Grover, A., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2018: 3077–84
  • Approximate Inference via Weighted Rademacher Complexity Kuck, J., Sabharwal, A., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2018: 6376–83
  • Deterministic Policy Optimization by Combining Pathwise and Score Function Estimators for Discrete Action Spaces Levy, D., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2018: 3474–81
  • Constructing Unrestricted Adversarial Examples with Generative Models Song, Y., Shu, R., Kushman, N., Ermon, S., Bengio, S., Wallach, H., Larochelle, H., Grauman, K., CesaBianchi, N., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2018
  • Multi-Agent Generative Adversarial Imitation Learning Song, J., Ren, H., Sadigh, D., Ermon, S., Bengio, S., Wallach, H., Larochelle, H., Grauman, K., CesaBianchi, N., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2018
  • Streamlining Variational Inference for Constraint Satisfaction Problems Grover, A., Achim, T., Ermon, S., Bengio, S., Wallach, H., Larochelle, H., Grauman, K., CesaBianchi, N., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2018
  • Coupling between oxygen redox and cation migration explains unusual electrochemistry in lithium-rich layered oxides NATURE COMMUNICATIONS Gent, W. E., Lim, K., Liang, Y., Li, Q., Barnes, T., Ahn, S., Stone, K. H., McIntire, M., Hong, J., Song, J., Li, Y., Mehta, A., Ermon, S., Tyliszczak, T., Kilcoyne, D., Vine, D., Park, J., Doo, S., Toney, M. F., Yang, W., Prendergast, D., Chueh, W. C. 2017; 8
  • A Survey on Behavior Recognition Using WiFi Channel State Information IEEE COMMUNICATIONS MAGAZINE Yousefi, S., Narui, H., Dayal, S., Ermon, S., Valaee, S. 2017; 55 (10): 98-104
  • Autotuning Stencil Computations with Structural Ordinal Regression Learning Cosenza, B., Durillo, J. J., Ermon, S., Juurlink, B., IEEE IEEE. 2017: 287–96
  • Label-Free Supervision of Neural Networks with Physics and Domain Knowledge Stewart, R., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2017: 2576-2582
  • Deep Gaussian Process for Crop Yield Prediction Based on Remote Sensing Data You, J., Li, X., Low, M., Lobell, D., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2017: 4559-4565
  • General Bounds on Satisfiability Thresholds for Random CSPs via Fourier Analysis Wei, C., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2017: 3958-3965
  • Estimating Uncertainty Online Against an Adversary Kuleshov, V., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2017: 2110-2116
  • Deep Hybrid Models: Bridging Discriminative and Generative Approaches Kuleshov, V., Ermon, S., AUAI AUAI PRESS. 2017
  • Fast Amortized Inference and Learning in Log-linear Models with Randomly Perturbed Nearest Neighbor Search Mussmann, S., Levy, D., Ermon, S., AUAI AUAI PRESS. 2017
  • A-NICE-MC: Adversarial Training for MCMC Song, J., Zhao, S., Ermon, S., Guyon, Luxburg, U. V., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2017
  • InfoGAIL: Interpretable Imitation Learning from Visual Demonstrations Li, Y., Song, J., Ermon, S., Guyon, Luxburg, U. V., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2017
  • Coupling between oxygen redox and cation migration explains unusual electrochemistry in lithium-rich layered oxides. Nature communications Gent, W. E., Lim, K. n., Liang, Y. n., Li, Q. n., Barnes, T. n., Ahn, S. J., Stone, K. H., McIntire, M. n., Hong, J. n., Song, J. H., Li, Y. n., Mehta, A. n., Ermon, S. n., Tyliszczak, T. n., Kilcoyne, D. n., Vine, D. n., Park, J. H., Doo, S. K., Toney, M. F., Yang, W. n., Prendergast, D. n., Chueh, W. C. 2017; 8 (1): 2091

    Abstract

    Lithium-rich layered transition metal oxide positive electrodes offer access to anion redox at high potentials, thereby promising high energy densities for lithium-ion batteries. However, anion redox is also associated with several unfavorable electrochemical properties, such as open-circuit voltage hysteresis. Here we reveal that in Li1.17-x Ni0.21Co0.08Mn0.54O2, these properties arise from a strong coupling between anion redox and cation migration. We combine various X-ray spectroscopic, microscopic, and structural probes to show that partially reversible transition metal migration decreases the potential of the bulk oxygen redox couple by > 1 V, leading to a reordering in the anionic and cationic redox potentials during cycling. First principles calculations show that this is due to the drastic change in the local oxygen coordination environments associated with the transition metal migration. We propose that this mechanism is involved in stabilizing the oxygen redox couple, which we observe spectroscopically to persist for 500 charge/discharge cycles.

    View details for PubMedID 29233965

    View details for PubMedCentralID PMC5727078

  • Monitoring Ethiopian Wheat Fungus with Satellite Imagery and Deep Feature Learning Pryzant, R., Ermon, S., Lobell, D., IEEE IEEE. 2017: 1524–32
  • Unsupervised Data Mining in nanoscale X-ray Spectro-Microscopic Study of NdFeB Magnet SCIENTIFIC REPORTS Duan, X., Yang, F., Antono, E., Yang, W., Pianetta, P., Ermon, S., Mehta, A., Liu, Y. 2016; 6

    Abstract

    Novel developments in X-ray based spectro-microscopic characterization techniques have increased the rate of acquisition of spatially resolved spectroscopic data by several orders of magnitude over what was possible a few years ago. This accelerated data acquisition, with high spatial resolution at nanoscale and sensitivity to subtle differences in chemistry and atomic structure, provides a unique opportunity to investigate hierarchically complex and structurally heterogeneous systems found in functional devices and materials systems. However, handling and analyzing the large volume data generated poses significant challenges. Here we apply an unsupervised data-mining algorithm known as DBSCAN to study a rare-earth element based permanent magnet material, Nd2Fe14B. We are able to reduce a large spectro-microscopic dataset of over 300,000 spectra to 3, preserving much of the underlying information. Scientists can easily and quickly analyze in detail three characteristic spectra. Our approach can rapidly provide a concise representation of a large and complex dataset to materials scientists and chemists. For example, it shows that the surface of common Nd2Fe14B magnet is chemically and structurally very different from the bulk, suggesting a possible surface alteration effect possibly due to the corrosion, which could affect the material's overall properties.

    View details for DOI 10.1038/srep34406

    View details for Web of Science ID 000384188300001

    View details for PubMedID 27680388

    View details for PubMedCentralID PMC5041149

  • Combining satellite imagery and machine learning to predict poverty. Science Jean, N., Burke, M., Xie, M., Davis, W. M., Lobell, D. B., Ermon, S. 2016; 353 (6301): 790-794

    Abstract

    Reliable data on economic livelihoods remain scarce in the developing world, hampering efforts to study these outcomes and to design policies that improve them. Here we demonstrate an accurate, inexpensive, and scalable method for estimating consumption expenditure and asset wealth from high-resolution satellite imagery. Using survey and satellite data from five African countries--Nigeria, Tanzania, Uganda, Malawi, and Rwanda--we show how a convolutional neural network can be trained to identify image features that can explain up to 75% of the variation in local-level economic outcomes. Our method, which requires only publicly available data, could transform efforts to track and target poverty in developing countries. It also demonstrates how powerful machine learning techniques can be applied in a setting with limited training data, suggesting broad potential application across many scientific domains.

    View details for DOI 10.1126/science.aaf7894

    View details for PubMedID 27540167

  • Transfer Learning from Deep Features for Remote Sensing and Poverty Mapping Xie, M., Jean, N., Burke, M., Lobell, D., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2016: 3929–35
  • Generative Adversarial Imitation Learning Ho, J., Ermon, S., Lee, D. D., Sugiyama, M., Luxburg, U. V., Guyon, Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2016
  • Probabilistic Inference by Hashing and Optimization PERTURBATIONS, OPTIMIZATION, AND STATISTICS Ermon, S., Hazan, T., Papandreou, G., Tarlow, D. 2016: 265-288
  • Tight Variational Bounds via Random Projections and I-Projections Hsu, L., Achim, T., Ermon, S., Gretton, A., Robert, C. C. MICROTOME PUBLISHING. 2016: 1087-1095
  • Closing the Gap Between Short and Long XORs for Model Counting Zhao, S., Chaturapruek, S., Sabharwal, A., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2016: 3322-3328
  • Exact Sampling with Integer Linear Programs and Random Perturbations Kim, C., Sabharwal, A., Ermon, S., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2016: 3248-3254
  • Adaptive Concentration Inequalities for Sequential Decision Problems Zhao, S., Zhou, E., Sabharwal, A., Ermon, S., Lee, D. D., Sugiyama, M., Luxburg, U. V., Guyon, Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2016
  • Solving Marginal MAP Problems with NP Oracles and Parity Constraints Xue, Y., Li, Z., Ermon, S., Gomes, C. P., Selman, B., Lee, D. D., Sugiyama, M., Luxburg, U. V., Guyon, Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2016
  • Variational Bayes on Monte Carlo Steroids Grover, A., Ermon, S., Lee, D. D., Sugiyama, M., Luxburg, U. V., Guyon, Garnett, R. NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2016
  • Learning Large-Scale Dynamic Discrete Choice Models of Spatio-Temporal Preferences with Application to Migratory Pastoralism in East Africa Ermon, S., Xue, Y., Toth, R., Dilkina, B., Bernstein, R., Damoulas, T., Clark, P., DeGloria, S., Mude, A., Barrett, C., Gomes, C. P., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2015: 644-650
  • Uncovering Hidden Structure through Parallel Problem Decomposition for the Set Basis Problem: Application to Materials Discovery Xue, Y., Ermon, S., Gomes, C. P., Selman, B., Yang, Q., Wooldridge, M. IJCAI-INT JOINT CONF ARTIF INTELL. 2015: 146-154
  • Importance Sampling over Sets: A New Probabilistic Inference Scheme Hadjis, S., Ermon, S., Meila, M., Heskes, T. AUAI PRESS. 2015: 355-364
  • Pattern Decomposition with Complex Combinatorial Constraints: Application to Materials Discovery Ermon, S., Le Bras, R., Suram, S. K., Gregoire, J. M., Gomes, C. P., Selman, B., van Dover, R. B., AAAI ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2015: 636-643