• No results found

Dynamical systems and learning

Learning Symbolic Representations of Hybrid Dynamical Systems

Learning Symbolic Representations of Hybrid Dynamical Systems

... arbitrary systems often require vast numbers of parameters, which obfuscates the interpretability of the inferred model (Breiman, ...hybrid dynamical systems which can be easily and naturally ...

34

Nonparametric Bayesian Learning of Switching Linear Dynamical Systems

Nonparametric Bayesian Learning of Switching Linear Dynamical Systems

... Linear dynamical systems (LDSs) are useful in describing dynamical phenomena as diverse as hu- man motion [9], financial time-series [4], maneuvering targets [6, 10], and the dance of honey bees ...

10

Deep Learning via Dynamical Systems: An Approximation Perspective

Deep Learning via Dynamical Systems: An Approximation Perspective

... deep learning, and more importantly, shed light on the role of composition on function approximation and ...the dynamical systems approach led to much progress in terms of novel algorithms [37, 53], ...

46

Causal learning for partially observed stochastic dynamical systems

Causal learning for partially observed stochastic dynamical systems

... of dynamical systems have causal interpretations that support reasoning about the consequences of interventions, suita- bly ...causal learning based on this notion of ...complete learning ...

17

Learning interpretable continuous-time models of latent stochastic dynamical systems

Learning interpretable continuous-time models of latent stochastic dynamical systems

... stochastic dynamical systems Lea Duncker 1 Gerg˝o Bohner 1 Julien Boussard 2 Maneesh Sahani 1 Abstract We develop an approach to learn an interpretable semi-parametric model of a latent continuous- time ...

9

On-Line Learning of Linear Dynamical Systems: Exponential Forgetting in Kalman Filters

On-Line Learning of Linear Dynamical Systems: Exponential Forgetting in Kalman Filters

... Linear Dynamical Systems (LDS) are a key standard tool in modeling and forecasting time series, with an exceedingly large number of ...of learning system parameters, via ex- pectation maximization ...

8

Learning and policy search in stochastic dynamical systems with Bayesian neural networks

Learning and policy search in stochastic dynamical systems with Bayesian neural networks

... stochastic dynamical systems using model-based reinforcement ...After learning the dynamics, our BNNs are then fed into an algorithm that performs random roll-outs and uses stochastic optimization ...

14

Dynamical systems for learning and balancing

Dynamical systems for learning and balancing

... As a starting point in our research it is proposed that familiar least squares techniques be used to learn functional representations of parameters of an ARMAX sy[r] ...

137

A scaled gradient projection method for Bayesian learning in dynamical systems

A scaled gradient projection method for Bayesian learning in dynamical systems

... In this paper we show that, with a suitable choice of the scaling matrix and a careful implementation, the scaled gradient projection method applied to the impulse response estimation pr[r] ...

22

Learning Stable Linear Dynamical Systems with the Weighted Least Square Method

Learning Stable Linear Dynamical Systems with the Weighted Least Square Method

... ical Systems (LDSs) from time series with the least-square method, where the stability of the sys- tem is not naturally ...for learning stable sys- tems by enforcing stability directly on the least- square ...

7

Block sparse linear models for learning structured dynamical systems in aeronautics

Block sparse linear models for learning structured dynamical systems in aeronautics

... We developed in this paper a new method for aircraft dynamics identification based on a multi-task learning framework. As it has been designed for trajec- tory optimization purposes, it delivers interpretable ...

18

Information transfer in dynamical systems

Information transfer in dynamical systems

... offers a particular advantage. In particular, robust optimization viewpoint provides a systematic way of determining the regularization parameter which often is a tuning parameter. On the other hand, the optimization ...

178

On dynamical systems

On dynamical systems

... We express the quotient space in terms of a non transitive subshift of finite type, give a necessary and sufficient condition for the existence of a local produc[r] ...

113

Dynamical Causal Learning

Dynamical Causal Learning

... Real causal learning tasks often involve uncertainty about both structure and parameters. Thus, even when a task demands ratings of causal strength, the structural uncertainty should still be taken into account; ...

8

Artificial Biochemical Networks : Evolving Dynamical Systems to Control Dynamical Systems

Artificial Biochemical Networks : Evolving Dynamical Systems to Control Dynamical Systems

... computational dynamical systems include reaction-diffusion based controllers [29] and echo state networks ...designing dynamical systems by hand, the use of evolutionary algorithms provides a ...

23

Equivariant dynamical systems

Equivariant dynamical systems

... theory for closed orbits of equivariant vector fields.. Essentially we pull back the flow in a tubular nbi.[r] ...

218

On Adjoint Dynamical Systems

On Adjoint Dynamical Systems

... The above results are referring only to generating supercategories S of adjoint dynamical systems, while Proposition 3 in I is referring to the "state space" X~ (o[r] ...

12

MODELLING OF DYNAMICAL SYSTEMS

MODELLING OF DYNAMICAL SYSTEMS

... the dynamical systems is in the focus of attention of scientists since many ...the dynamical systems. Dynamical systems can be technical ...electrical systems, mechanical ...

20

Dynamical Systems in Neuroscience

Dynamical Systems in Neuroscience

... Surprisingly, this chapter turned out to be quite different from chapter 9 (“Weakly Connected Oscillators”) of the book Weakly Connected Neural Networks by Hoppen- steadt and Izhikevich (1997) and from the book ...

210

Dynamical Systems on Bundles

Dynamical Systems on Bundles

... mechanical systems, are structured on the horizontal and the vertical distributions of tangent and cotangent ...Lagrangian dynamical systems are ...

11

Show all 10000 documents...

Related subjects