Stochastic analyses of the dynamics of generalized Little-Hopfield-Hemmen type neural networks

Stochastic analyses are conducted of model neural networks of the generalized Little-Hopfield-Hemmen type, in which the synaptic connections with linearly embeddedp sets of patterns are free of symmetric ones, and a Glauber dynamics of a Markovian type is assumed. Two kinds of approaches are taken to study the stochastic dynamical behavior of the network system. First, by developing the method of the nonlinear master equation in the thermodynamic limitN, an exact self-consistent equation is derived for the time evolultion of the pattern overlaps which play the role of the order parameters of the system. The self-consistent equation is shown to describe almost completely the macroscopic dynamical behavior of the network system. Second, conducting the system-size expansion of the master equation for theN-body probability distribution of the Glauber dynamics makes it possible to analyze the fluctuations. In the course of the analysis, the self-consistent equation for the pattern overlaps is derived again. The main result of the rigorous fluctuation analysis is that as far as the fluctuations are concerned, the time course of the pattern overlap fluctuations behaves independently of the fluctuations in the remaining modes of the system's macrovariables, in accordance with the self-determining property of the macroscopic motion of the pattern overlaps for neural networks with linear synaptic couplings.     

Master equation techniques for exciton motion,relaxation, capture,and annihilation

Techniques developed recently for the study of exciton dynamics in molecular solids are discussed. They include master equation methods for the analysis of prerelaxation energy transfer, the generalized master equation approach to coherence in exciton motion, and the defect technique as applied to exciton trapping and annihilation.     

Non-Markovian dynamics of a qubit in a reservoir: different solutions of non-Markovian master equation

We present a non-Markovian master equation for a qubit interacting with a general reservoir, which is derived according to the Nakajima-Zwanzig and the time convolutionless projection operator technique. The non-Markovian solutions and Markovian solution of dynamical decay of a qubit are compared. The results indicate the validity of non-Markovian approach in different coupling regimes and also show that the Markovian master equation may not precisely describe the dynamics of an open quantum system in some situation. The non-Markovian solutions may be effective for many qubits independently interacting with the heated reservoirs.     

Geometry and nonlinear evolution equations

We briefly review the nonlinear dynamics of diverse physical systems which can be described in terms of moving curves and surfaces. The interesting connections that exist between the underlying differential geometry of these systems and the corresponding nonlinear partial differential equations are highlighted by considering classic examples such as the motion of a vortex filament in a fluid and the dynamics of a spin chain. The association of the dynamics of a non-stretching curve with a hierarchy of completely integrable soliton-supporting equations is discussed. The application of the surface embeddability approach is shown to be useful in obtaining such connections as well as exact solutions of some nonlinear systems such as the Belavin-Polyakov equation and the inhomogeneous Heisenberg chain.     

Nonlinear relaxation and fluctuations of damped quantum systems

The paper reexamines the treatment of irreversible quantum systems by master equations. Shortcomings of the conventional theory of quantum Markov processes pointed out by Talkner are analyzed. It is shown that a frequently used quantum regression hypothesis is not correct, in general. A new generalized master equation determining the relaxation to equilibrium is derived by means of time-dependent projection operator techniques. It is shown that this master equation also determines the time evolution of equilibrium correlations and response functions. The Markovian approximation is discussed, and a new type of Markovian limit, the Brownian motion limit, is introduced besides the weak coupling limit. The shortcomings of the conventional approach are resolved by deriving new formulae for the time evolution of the correlation and response functions of a quantum Markov process. The symmetries of the process are emphasized, and it is shown how the fluctuation-dissipation theorem and the detailed balance symmetry emerge from the master equation approach.     

Stochastic exponential stability of the delayed reaction-diffusion recurrent neural networks with Markovian jumping parameters

Some criteria for the global stochastic exponential stability of the delayed reaction-diffusion recurrent neural networks with Markovian jumping parameters are presented. The jumping parameters considered here are generated from a continuous-time discrete-state homogeneous Markov process, which are governed by a Markov process with discrete and finite state space. By employing a new Lyapunov-Krasovskii functional, a linear matrix inequality (LMI) approach is developed to establish some easy-to-test criteria of global exponential stability in the mean square for the stochastic neural networks. The criteria are computationally efficient, since they are in the forms of some linear matrix inequalities.     

Escape rates in periodically driven Markov processes

We present an approximate analytical expression for escape rates of time-dependent driven stochastic processes with an absorbing boundary such as the driven leaky integrate-and-fire model for neural spiking. The novel approximation is based on a discrete state Markovian modeling of the full long-time dynamics with time-dependent rates. It is valid in a wide parameter regime beyond the restraining limits of weak driving (linear response) and/or weak noise. The scheme is carefully tested and yields excellent agreement with three different numerical methods based on the Langevin equation, the Fokker–Planck equation and an integral equation.     

Assessing non-Markovian quantum dynamics

We investigate what a snapshot of a quantum evolution--a quantum channel reflecting open system dynamics--reveals about the underlying continuous time evolution. Remarkably, from such a snapshot, and without imposing additional assumptions, it can be decided whether or not a channel is consistent with a time (in)dependent Markovian evolution, for which we provide computable necessary and sufficient criteria. Based on these, a computable measure of "Markovianity" is introduced. We discuss how the consistency with Markovian dynamics can be checked in quantum process tomography. The results also clarify the geometry of the set of quantum channels with respect to being solutions of time (in)dependent master equations.     

Optimal Population Coding for Dynamic Input by Nonequilibrium Networks

The efficient coding hypothesis states that neural response should maximize its information about the external input. Theoretical studies focus on optimal response in single neuron and population code in networks with weak pairwise interactions. However, more biological settings with asymmetric connectivity and the encoding for dynamical stimuli have not been well-characterized. Here, we study the collective response in a kinetic Ising model that encodes the dynamic input. We apply gradient-based method and mean-field approximation to reconstruct networks given the neural code that encodes dynamic input patterns. We measure network asymmetry, decoding performance, and entropy production from networks that generate optimal population code. We analyze how stimulus correlation, time scale, and reliability of the network affect optimal encoding networks. Specifically, we find network dynamics altered by statistics of the dynamic input, identify stimulus encoding strategies, and show optimal effective temperature in the asymmetric networks. We further discuss how this approach connects to the Bayesian framework and continuous recurrent neural networks. Together, these results bridge concepts of nonequilibrium physics with the analyses of dynamics and coding in networks.     

Micromaser line broadening without photon exchange

We perform a calculation of the linewidth of a micromaser, using the master equation and the quantum regression approach. A `dephasing' contribution is identified from pumping processes that conserve the photon number and do not appear in the photon statistics. We work out examples for a single-atom maser with a precisely controlled coupling and for a laser where the interaction time is broadly distributed. In the latter case, we also assess the convergence of a recently developed uniform Lindblad approximation to the master equation; it is relatively slow.     

Significance of Updating Schemes in Computational Models: Dynamics of Neutral Networks

A Hopfield neural network was constructed with relevance to protein dynamics. The dynamics of this network was analyzed by determining the distribution of first passage times between memories and its dependence on temperature. The distribution depended on the updating scheme. This illustrates the importance of choosing an updating scheme that leads to physically meaningful results in computational models of dynamic processes, such as in neural networks or molecular dynamics.     

Robust stability analysis for Markovian jumping stochastic neural networks with mode-dependent time-varying interval delay and multiplicative noise

This paper is concerned with the problem of robust stability for a class of Markovian jumping stochastic neural networks (MJSNNs) subject to mode-dependent time-varying interval delay and state-multiplicative noise. Based on the Lyapunov--Krasovskii functional and a stochastic analysis approach, some new delay-dependent sufficient conditions are obtained in the linear matrix inequality (LMI) format such that delayed MJSNNs are globally asymptotically stable in the mean-square sense for all admissible uncertainties. An important feature of the results is that the stability criteria are dependent on not only the lower bound and upper bound of delay for all modes but also the covariance matrix consisting of the correlation coefficient. Numerical examples are given to illustrate the effectiveness.     

Influence of Lamb Shift Parameter on Dissipative Dynamics of the Phase Damped Jaynes–Cummings Model with Gravity Under Markovian Approximation

In this paper, we study the dissipative dynamics of the phase damped Jaynes–Cummings model with gravity under Markovian approximation in the presence of the Lamb shift parameter. The model consists of a moving two-level atom simultaneously exposed to the gravitational field and a single-mode traveling radiation field in the presence of a phase damping mechanism. We first present the master equation for the reduced density operator of the system under Markovian approximation in terms of a Hamiltonian describing the atom-field interaction with gravity in the presence of Lamb-shift parameter. Then, by making use of the super-operator technique, we obtain an exact solution of the master equation. Assuming that initially the radiation field is prepared in a Glauber coherent state and the two-level atom is in the excited state, we investigate the influence of Lamb shift parameter on the temporal evolution of collapses and revivals of the atomic population inversion, atomic dipole squeezing and atomic momentum diffusion in the presence of phase damping.     

耗散子运动方程组的可控截断

为耗散子运动方程组提出精度可控的预截断方案来模拟和研究非微扰的量子耗散动力学．发展了具有类似Caldeira-Leggett主方程形式、在终止阶层采取马尔科夫和高温近似方法的截断方案,根据马尔科夫性质的分析给出确定终止层数的精度判据,并通过电荷转移体系的动力学模拟进行了检验．     

Quantum Evolution beyond the Markovian Semigroup — Generalizing the Stenholm–Barnett Approach

We provide conditions for the memory kernel governing the time-nonlocal quantum master equation which guarantee that the corresponding dynamical map is completely positive and trace-preserving. This approach gives rise to the new parametrization of dynamical maps in terms of two completely positive maps – so-called legitimate pair. In fact, these new parameterizations are a natural generalization of Markovian semigroup. Interestingly our class contains recently studied models like semi-Markov evolution and collision models.     

从单分子到凝聚相水体系的理论描述：势能面和分子动力学最新进展

本文回顾了近十年来水体系的势能面与分子动力学理论研究的最新进展,包括水分子参与的气相反应,固体表面上的吸附与解离动力学,以及从团簇到凝聚相水的结构、振动光谱与统计力学模拟. 近年来再次发展起来的机器学习技术,例如结合置换不变多项式的神经网络,或结合基本不变量的神经网络,已被成熟应用于气相与固体表面体系的高精度势能面构造中. 对于团簇甚至凝聚相水体系,原子中心神经网络方法或基于核的高斯过程方法应用更为广泛. 此外,在多体展开框架下,在气相体系中发展起来的的方法也组成了高维度体系势函数构造的高精度方案. 当前凝聚相水体系面临的主要问题是高精度从头算数据集的积累,兼顾计算精度与效率的双杂化密度泛函是一种可能的解决方案. 在动力学理论方面,无论是化学反应截面计算还是振动光谱模拟,往往需要合理描述水分子中氢原子的量子效应,才能得到较为可靠的理论计算结果. 量子波包动力学方法已经在气相反应机理研究方面有深入的应用,也在包含数个水分子的团簇振动分析中有初步应用. 基于路径积分的分子动力学方法正在较大水团簇以及凝聚相水的结构与谱学模拟方面发挥重要作用.     

Influence of a Classical Homogeneous Gravitational Field on Dissipative Dynamics of the Phase Damped Jaynes–Cummings Model under the Markovian Approximation

In this paper, we study the dissipative dynamics of the phase damped Jaynes–Cummings model under the Markovian approximation in the presence of a classical homogeneous gravitational field. The model consists of a moving two-level atom simultaneously exposed to the gravitational field and a single-mode traveling radiation field in the presence of a phase damping mechanism. We first present the master equation for the reduced density operator of the system under the Markovian approximation in terms of a Hamiltonian describing the atom-field interaction in the presence of a homogeneous gravitational field. Then, by making use of the super-operator technique, we obtain an exact solution of the master equation. Assuming that initially the radiation field is prepared in a Glauber coherent state and the two-level atom is in the excited state, we investigate the influence of gravity on the temporal evolution of collapses and revivals of the atomic population inversion, atomic dipole squeezing, atomic momentum diffusion, photon counting statistics and quadrature squeezing of the radiation field in the presence of phase damping.     

Path summation formulation of the master equation

Markovian dynamics, modeled by the kinetic master equation, has wide ranging applications in chemistry, physics, and biology. We derive an exact expression for the probability of a Markovian path in discrete state space for an arbitrary number of states and path length. The total probability of paths repeatedly visiting a set of states can be explicitly summed. The transition probability between states can be expressed as a sum over all possible paths connecting the states. The derived path probabilities satisfy the fluctuation theorem. The paths can be the starting point for a path space Monte Carlo procedure which can serve as an alternative algorithm to analyze pathways in a complex reaction network.