54 Hits in 5.8 sec

Model based control of dynamic atomic force microscope

Chibum Lee, Srinivasa M. Salapaka
2015 Review of Scientific Instruments  
Obeying the detailed-balance principle, the system equilibrates to thermal and chemical equilibrium with dynamical fluctuations on the fields, generated dynamically by the discrete interactions.  ...  We generalize an existing framework [Bashan , 2008] for adaptive allocation of sensing resources to the dynamic case, accounting for time-varying target behavior such as transitions to neighboring cells  ...  (OMP) and Basis Pursuit (BP) are two well-known recovery algorithms in compressed sensing.  ... 
doi:10.1063/1.4917301 pmid:25933864 fatcat:vdcat7sgdncujkkgyuklvbwgwe

Sensing Cox Processes via Posterior Sampling and Positive Bases [article]

Mojmír Mutný, Andreas Krause
2022 arXiv   pre-print
Our adaptive sensing algorithms use Langevin dynamics and are based on posterior sampling (Cox-Thompson) and top-two posterior sampling (Top2) principles.  ...  With latter, the difference between samples serves as a surrogate to the uncertainty.  ...  Langevin dynamics applies.  ... 
arXiv:2110.11181v2 fatcat:bq4qo7ss6be5hazfosfkytsryu

Denoising Score-Matching for Uncertainty Quantification in Inverse Problems [article]

Zaccharie Ramzi, Benjamin Remy, Francois Lanusse, Jean-Luc Starck, Philippe Ciuciu
2020 arXiv   pre-print
In this work, we propose a generic Bayesian framework forsolving inverse problems, in which we limit the use of deep neural networks tolearning a prior distribution on the signals to recover.  ...  We adopt recent denoisingscore matching techniques to learn this prior from data, and subsequently use it aspart of an annealed Hamiltonian Monte-Carlo scheme to sample the full posteriorof image inverse  ...  dynamics.  ... 
arXiv:2011.08698v1 fatcat:ukjxwlxnffbjvhn5vq6mf3u3ui

Molecular Dynamics: Survey of Methods for Simulating the Activity of Proteins

Stewart A. Adcock, J. Andrew McCammon
2006 Chemical Reviews  
., the progress of simulated structure with respect to time) generally provides data only at the level of atomic positions, velocities, and single-point energies.  ...  The majority of important dynamics methodologies are highly dependent upon the availability of a suitable potential-energy function to describe the energy landscape of the system with respect to the aforementioned  ...  Langevin Dynamics Langevin dynamics incorporates stochastic terms to approximate the effects of degrees of freedom that are neglected in the simulation.  ... 
doi:10.1021/cr040426m pmid:16683746 pmcid:PMC2547409 fatcat:xgwmyq7covhp7otzqqqfj23xbm

Big Learning with Bayesian Methods [article]

Jun Zhu, Jianfei Chen, Wenbo Hu, Bo Zhang
2017 arXiv   pre-print
, regularized Bayesian inference for improving the flexibility via posterior regularization, and scalable algorithms and systems based on stochastic subsampling and distributed computing for dealing with  ...  Bayesian methods represent one important class of statistic methods for machine learning, with substantial recent developments on adaptive, flexible and scalable Bayesian learning.  ...  Stochastic Gradient: The idea of using gradient information to improve the mixing rates has been systematically studied in various MC methods, including Langevin dynamics and Hamiltanian dynamics [136  ... 
arXiv:1411.6370v2 fatcat:zmxse4kkqjgffkricevyumaoiu

A Modularized Efficient Framework for Non-Markov Time Series Estimation

Gabriel Schamberg, Demba Ba, Todd P. Coleman
2018 IEEE Transactions on Signal Processing  
dynamic analyses of learning with neural spiking and behavioral observations.  ...  and subsequently "averaging" them in an appropriate sense using a Kalman smoother.  ...  on dynamics, and "averaging" the two in the appropriate sense.  ... 
doi:10.1109/tsp.2018.2793870 fatcat:wfapnqqe2rcvnnsxnhhlciszwe

Stein's Method Meets Statistics: A Review of Some Recent Developments [article]

Andreas Anastasiou, Alessandro Barp, François-Xavier Briol, Bruno Ebner, Robert E. Gaunt, Fatemeh Ghaderinezhad, Jackson Gorham, Arthur Gretton, Christophe Ley, Qiang Liu, Lester Mackey, Chris. J. Oates (+2 others)
2021 arXiv   pre-print
The topics we discuss include: explicit error bounds for asymptotic approximations of estimators and test statistics, a measure of prior sensitivity in Bayesian statistics, tools to benchmark and compare  ...  Stein Operators via the Generator Approach We first describe the generator approach, which we present for a given target P on X = R d .  ...  If the mean is the parameter of interest, Ley et al. (2017a) compare a normal N(µ, δ 2 ) prior for the location parameter with a uniform prior (we note that a normal prior is the conjugate prior in this  ... 
arXiv:2105.03481v1 fatcat:v3g4gxzxkja5hm6vjsqenip24e

Deep Learning is Singular, and That's Good [article]

Daniel Murfet, Susan Wei, Mingming Gong, Hui Li, Jesse Gell-Redman, Thomas Quella
2020 arXiv   pre-print
In singular models, the optimal set of parameters forms an analytic set with singularities and classical statistical inference cannot be applied to such models.  ...  Via a mix of theory and experiment, we present an invitation to singular learning theory as a vehicle for understanding deep learning and suggest important future work to make singular learning theory  ...  Stronger generalization bounds for deep nets via a compression approach. In 35th International Conference on Machine Learning, ICML 2018, 2018. Yuanzhi Li and Yingyu Liang.  ... 
arXiv:2010.11560v1 fatcat:chxkaqevf5guzburlgp4baag2q

Proceedings of the second "international Traveling Workshop on Interactions between Sparse models and Technology" (iTWIST'14) [article]

L. Jacques, C. De Vleeschouwer, Y. Boursier, P. Sudhakar, C. De Mol, A. Pizurica, S. Anthoine, P. Vandergheynst, P. Frossard, C. Bilen, S. Kitic, N. Bertin, R. Gribonval, N. Boumal (+51 others)
2014 arXiv   pre-print
of the "sparsity paradigm": Sparsity-driven data sensing and processing; Union of low dimensional subspaces; Beyond linear and convex inverse problem; Matrix/manifold/graph sensing/processing; Blind inverse  ...  iTWIST'14 has gathered about 70 international participants and has featured 9 invited talks, 10 oral presentations, and 14 posters on the following themes, all related to the theory, application and generalization  ...  There, the authors study 2 -stability for this class of decomposable norms with a general sufficiently smooth data fidelity.  ... 
arXiv:1410.0719v2 fatcat:4y3drgk3ujh5hopfn2p2runlzu

Bayesian inference via sparse Hamiltonian flows [article]

Naitong Chen, Zuheng Xu, Trevor Campbell
2022 arXiv   pre-print
Real and synthetic experiments demonstrate that sparse Hamiltonian flows provide accurate posterior approximations with significantly reduced runtime compared with competing dynamical-system-based inference  ...  Theoretical results show that the method enables an exponential compression of the dataset in a representative model, and that the quasi-refreshment steps reduce the KL divergence to the target.  ...  In a Bayesian inference problem with i.i.d. data, π 0 is the prior density, the f n are the log-likelihood terms for N data points, and the normalization constant is in general not known.  ... 
arXiv:2203.05723v1 fatcat:3cyhmuw2xffdzbgmlxak5mckua

Elements of naturality in dynamical simulation frameworks for Hamiltonian, thermostatic, and Lindbladian flows on classical and quantum state-spaces [article]

John A. Sidles, Joseph L. Garbini, Jonathan P. Jacky, Rico A. R. Picone, Scott A. Harsila
2010 arXiv   pre-print
Both classical and quantum examples are presented, including dynamic nuclear polarization (DNP).  ...  specifying dynamical potentials that are physically natural; in each respect explicit criteria are given for "naturality."  ...  all three) act generically to compress trajectories-both classical and quantum-onto state-spaces of lower dimension.  ... 
arXiv:1007.1958v1 fatcat:7oeocaykvfgl5h5ptvk6ukmykm

Bayesian imaging using Plug Play priors: when Langevin meets Tweedie [article]

Rémi Laumont, Valentin de Bortoli, Andrés Almansa, Julie Delon, Alain Durmus, Marcelo Pereyra
2022 arXiv   pre-print
To address these limitations, this paper develops theory, methods, and provably convergent algorithms for performing Bayesian inference with PnP priors.  ...  These methods derive Minimum Mean Square Error (MMSE) or Maximum A Posteriori (MAP) estimators for inverse problems in imaging by combining an explicit likelihood function with a prior that is implicitly  ...  The use of Plug & Play operators has also been investigated in the context of Approximate Message Passing (AMP) computation methods (see [30] for an introduction to AMP focused on compressed sensing  ... 
arXiv:2103.04715v6 fatcat:4xryxmhvd5gvxa6xujnnp3d5w4

Generalized Energy Based Models [article]

Michael Arbel and Liang Zhou and Arthur Gretton
2021 arXiv   pre-print
These models combine two trained components: a base distribution (generally an implicit model), which can learn the support of data with low intrinsic dimension in a high dimensional space; and an energy  ...  Samples from the posterior on the latent space of the trained model can be obtained via MCMC, thus finding regions in this space that produce better quality samples.  ...  LOGAN: Latent Optimisation for Generative Adversarial Networks. arXiv:1912.00953 [cs, stat]. arXiv: 1912.00953. Wu, Y., Rosca, M., and Lillicrap, T. (2019b). Deep compressed sensing.  ... 
arXiv:2003.05033v5 fatcat:7unhslzgrbdnbnypupsuxgoqqa

Practical recipes for the model order reduction, dynamical simulation and compressive sampling of large-scale open quantum systems

John A Sidles, Joseph L Garbini, Lee E Harrell, Alfred O Hero, Jonathan P Jacky, Joseph R Malcomb, Anthony G Norman, Austin M Williamson
2009 New Journal of Physics  
Single-spin detection by magnetic resonance force microscopy (MRFM) is simulated, and the data statistics are shown to be those of a random telegraph signal with additive white noise.  ...  Prior support was provided by the ARO under contact no. DAAD19-02-1-0344, by IBM under subcontract no. A0550287 through ARO contract no.  ...  Acknowledgments Author J A Sidles gratefully acknowledges the insights gained from discussions with Al Matsen-to whom this paper is dedicated-upon ab initio quantum chemistry, with Rick Matsen upon opportunities  ... 
doi:10.1088/1367-2630/11/6/065002 fatcat:vhkdy5gotvdz5lxfhfozpmhllm

Steps Towards a Theory of Visual Information: Active Perception, Signal-to-Symbol Conversion and the Interplay Between Sensing and Control [article]

Stefano Soatto
2017 arXiv   pre-print
This has consequences in the so-called "signal-to-symbol barrier" problem, as well as in the analysis and design of active sensing systems.  ...  It is shown that the "actionable information gap" between the two can be reduced by exercising control on the sensing process. Thus, senging, control and information are inextricably tied.  ...  provable guarantees.  ... 
arXiv:1110.2053v4 fatcat:utdycuug75drzkm2a4s74ozeg4
« Previous Showing results 1 — 15 out of 54 results