Agnostic Physics-Driven Deep Learning
- URL: http://arxiv.org/abs/2205.15021v1
- Date: Mon, 30 May 2022 12:02:53 GMT
- Title: Agnostic Physics-Driven Deep Learning
- Authors: Benjamin Scellier, Siddhartha Mishra, Yoshua Bengio, Yann Ollivier
- Abstract summary: This work establishes that a physical system can perform statistical gradient learning without gradient computations.
In Aeqprop, the specifics of the system do not have to be known: the procedure is based on external manipulations.
Aeqprop also establishes that in natural (bio)physical systems, genuine gradient-based statistical learning may result from generic, relatively simple mechanisms.
- Score: 82.89993762912795
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: This work establishes that a physical system can perform statistical learning
without gradient computations, via an Agnostic Equilibrium Propagation
(Aeqprop) procedure that combines energy minimization, homeostatic control, and
nudging towards the correct response. In Aeqprop, the specifics of the system
do not have to be known: the procedure is based only on external manipulations,
and produces a stochastic gradient descent without explicit gradient
computations. Thanks to nudging, the system performs a true, order-one gradient
step for each training sample, in contrast with order-zero methods like
reinforcement or evolutionary strategies, which rely on trial and error. This
procedure considerably widens the range of potential hardware for statistical
learning to any system with enough controllable parameters, even if the details
of the system are poorly known. Aeqprop also establishes that in natural
(bio)physical systems, genuine gradient-based statistical learning may result
from generic, relatively simple mechanisms, without backpropagation and its
requirement for analytic knowledge of partial derivatives.
Related papers
- Learning Controlled Stochastic Differential Equations [61.82896036131116]
This work proposes a novel method for estimating both drift and diffusion coefficients of continuous, multidimensional, nonlinear controlled differential equations with non-uniform diffusion.
We provide strong theoretical guarantees, including finite-sample bounds for (L2), (Linfty), and risk metrics, with learning rates adaptive to coefficients' regularity.
Our method is available as an open-source Python library.
arXiv Detail & Related papers (2024-11-04T11:09:58Z) - Modeling Unknown Stochastic Dynamical System Subject to External Excitation [4.357350642401934]
We present a numerical method for learning unknown nonautonomous dynamical system.
Our basic assumption is that the governing equations for the system are unavailable.
When a sufficient amount of such I/O data are available, our method is capable of learning the unknown dynamics.
arXiv Detail & Related papers (2024-06-22T06:21:44Z) - Hierarchical-Hyperplane Kernels for Actively Learning Gaussian Process
Models of Nonstationary Systems [5.1672267755831705]
We present a kernel family that incorporates a partitioning that is learnable via gradient-based methods.
We empirically demonstrate excellent performance on various active learning tasks.
arXiv Detail & Related papers (2023-03-17T14:50:51Z) - A Causality-Based Learning Approach for Discovering the Underlying
Dynamics of Complex Systems from Partial Observations with Stochastic
Parameterization [1.2882319878552302]
This paper develops a new iterative learning algorithm for complex turbulent systems with partial observations.
It alternates between identifying model structures, recovering unobserved variables, and estimating parameters.
Numerical experiments show that the new algorithm succeeds in identifying the model structure and providing suitable parameterizations for many complex nonlinear systems.
arXiv Detail & Related papers (2022-08-19T00:35:03Z) - Structure-Preserving Learning Using Gaussian Processes and Variational
Integrators [62.31425348954686]
We propose the combination of a variational integrator for the nominal dynamics of a mechanical system and learning residual dynamics with Gaussian process regression.
We extend our approach to systems with known kinematic constraints and provide formal bounds on the prediction uncertainty.
arXiv Detail & Related papers (2021-12-10T11:09:29Z) - Random features for adaptive nonlinear control and prediction [15.354147587211031]
We propose a tractable algorithm for both adaptive control and adaptive prediction.
We approximate the unknown dynamics with a finite expansion in $textitrandom$ basis functions.
Remarkably, our explicit bounds only depend $textitpolynomially$ on the underlying parameters of the system.
arXiv Detail & Related papers (2021-06-07T13:15:40Z) - Gradient Starvation: A Learning Proclivity in Neural Networks [97.02382916372594]
Gradient Starvation arises when cross-entropy loss is minimized by capturing only a subset of features relevant for the task.
This work provides a theoretical explanation for the emergence of such feature imbalance in neural networks.
arXiv Detail & Related papers (2020-11-18T18:52:08Z) - Active Learning for Nonlinear System Identification with Guarantees [102.43355665393067]
We study a class of nonlinear dynamical systems whose state transitions depend linearly on a known feature embedding of state-action pairs.
We propose an active learning approach that achieves this by repeating three steps: trajectory planning, trajectory tracking, and re-estimation of the system from all available data.
We show that our method estimates nonlinear dynamical systems at a parametric rate, similar to the statistical rate of standard linear regression.
arXiv Detail & Related papers (2020-06-18T04:54:11Z) - On dissipative symplectic integration with applications to
gradient-based optimization [77.34726150561087]
We propose a geometric framework in which discretizations can be realized systematically.
We show that a generalization of symplectic to nonconservative and in particular dissipative Hamiltonian systems is able to preserve rates of convergence up to a controlled error.
arXiv Detail & Related papers (2020-04-15T00:36:49Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.