检索结果-内蒙古大学图书馆

The Hierarchical Discrete pursuit Learning Automaton: A Novel Scheme With Fast Convergence and Epsilon-Optimality

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2024年第6期35卷 8278-8292页

作者： Omslandseter, Rebekka Olsson Jiao, Lei Zhang, Xuan Yazidi, Anis Oommen, B. John Univ Agder Dept Informat & Commun Technol N-4879 Grimstad Norway Norwegian Res Ctr NORCE N-4879 Grimstad Norway Oslo Metropolitan Univ Dept Comp Sci N-0160 Oslo Norway Carleton Univ Sch Comp Sci Ottawa ON K1S 5B6 Canada Northwest Univ TRADE Res Entity ZA-2520 Potchefstroom South Africa

Since the early 1960s, the paradigm of learning automata (LA) has experienced abundant interest. Arguably, it has also served as the foundation for the phenomenon and field of reinforcement learning (RL). Over the decades, new concepts and fundamental principles have been introduced to increase the LA's speed and accuracy. These include using probability updating functions, discretizing the probability space, and using the "pursuit" concept. Very recently, the concept of incorporating "structure" into the ordering of the LA's actions has improved both the speed and accuracy of the corresponding hierarchical machines, when the number of actions is large. This has led to the epsilon-optimal hierarchical continuous pursuit LA (HCPA). This article pioneers the inclusion of all the above-mentioned phenomena into a new single LA, leading to the novel hierarchical discretized pursuit LA (HDPA). Indeed, although the previously proposed HCPA is powerful, its speed has an impediment when any action probability is close to unity, because the updates of the components of the probability vector are correspondingly smaller when any action probability becomes closer to unity. We propose here, the novel HDPA, where we infuse the phenomenon of discretization into the action probability vector's updating functionality, and which is invoked recursively at every stage of the machine's hierarchical structure. This discretized functionality does not possess the same impediment, because discretization prohibits it. We demonstrate the HDPA's robustness and validity by formally proving the epsilon-optimality by utilizing the moderation property. We also invoke the submartingale characteristic at every level, to prove that the action probability of the optimal action converges to unity as time goes to infinity. Apart from the new machine being epsilon-optimal, the numerical results demonstrate that the number of iterations required for convergence is significantly reduced for the HDPA, when

关键词： Learning automata Automata Convergence Reinforcement learning Task analysis pursuit algorithms Markov processes Convergence analysis hierarchical discrete pursuit LA learning automata (LA) reinforcement learning (RL)

来源：评论

学校读者我要写书评

暂无评论

DISCRETIZED pursuit LEARNING AUTOMATA

引用

IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS 1990年第4期20卷 931-938页

作者： OOMMEN, BJ LANCTOT, JK School of Computer Science Carleton University Ottawa ONT Canada

The problem of a stochastic learning automaton interacting with an unknown random environment is considered. The fundamental problem is that of learning, through interaction, the best action allowed by the environment (i.e. the action that is rewarded optimally). By using running estimates of reward probabilities to learn the optimal action, an extremely efficient pursuit algorithm (PA), which is presently among the fastest algorithms known, was reported in earlier works. The improvements gained by rendering the PA discrete are investigated. This is done by restricting the probability of selecting an action to a finite and, hence, discrete subset of (0, 1). This improved scheme is proven to be epsilon -optimal in all stationary environments. Furthermore, the experimental results seem to indicate that the algorithm presented is faster than the fastest nonestimator learning automata reported to date, and also faster than the continuous pursuit automaton.

关键词： learning automata Fasting Action algorithms pursuit algorithms Nylons Learning Finite Environments

来源：评论

学校读者我要写书评

暂无评论

GOOD WEIGHTS AND HYPERBOLIC KERNELS FOR NEURAL NETWORKS, PROJECTION pursuit, AND PATTERN-CLASSIFICATION - FOURIER STRATEGIES FOR EXTRACTING INFORMATION FROM HIGH-DIMENSIONAL DATA

引用

IEEE TRANSACTIONS ON INFORMATION THEORY 1994年第2期40卷 439-454页

作者： JONES, LK Department of Mathematics Sciences University of Massachusetts Lowell Lowell MA USA

Fourier approximation and estimation of discriminant, regression, and density functions are considered. A preference order is established for the frequency weights in multiple Fourier expansions and the connection weights in single hidden-layer neural networks. These preferred weight vectors, called good weights (good lattice weights for estimation of periodic functions), are generalizations for arbitrary periods of the hyperbolic lattice points of Korobov and Hlawka associated with classes of smooth functions of periods one in each variable. Although previous results on approximation and quadrature are affinely invariant to the scale of the underlying periods, some of our results deal with optimization over finite sets and strongly depend on the choice of scale. It is shown how to count and generate good lattice weights. Finite sample bounds on mean integrated squared error are calculated for ridge estimates of periodic pattern class densities. The bounds are combined with a table of cardinalities of good lattice weight sets to furnish classifier design with prescribed class density estimation errors. Applications are presented for neural networks and projection pursuit. A hyperbolic kernel gradient transform is developed which automatically determines the training weights (projection directions). Its sampling properties are discussed. algorithms are presented for generating good weights for projection pursuit.

关键词： Kernel Neural networks Lattices Density functional theory Frequency Estimation error Sampling methods pursuit algorithms

来源：评论

学校读者我要写书评

暂无评论

Complex and Quaternionic Principal Component pursuit and Its Application to Audio Separation

引用

IEEE SIGNAL PROCESSING LETTERS 2016年第2期23卷 287-291页

作者： Chan, Tak-Shing T. Yang, Yi-Hsuan Acad Sinica Res Ctr Informat Technol Innovat Taipei 11564 Taiwan

Recently, the principal component pursuit has received increasing attention in signal processing research ranging from source separation to video surveillance. So far, all existing formulations are real-valued and lack the concept of phase, which is inherent in inputs such as complex spectrograms or color images. Thus, in this letter, we extend principal component pursuit to the complex and quaternionic cases to account for the missing phase information. Specifically, we present both complex and quaternionic proximity operators for the l(1)- and trace-norm regularizers. These operators can be used in conjunction with proximal minimization methods such as the inexact augmented Lagrange multiplier algorithm. The new algorithms are then applied to the singing voice separation problem, which aims to separate the singing voice from the instrumental accompaniment. Results on the iKala and MSD100 datasets confirmed the usefulness of phase information in principal component pursuit.

关键词： Principal component pursuit algorithms quaternions source separation

来源：评论

学校读者我要写书评

暂无评论

Randomized Simultaneous Hard Thresholding pursuit Algorithm for Multiple Measurement Vectors 14

Randomized Simultaneous Hard Thresholding Pursuit Algorithm ...

引用

14th IEEE International Conference on Signal Processing and Communications (SPCOM)

作者： Bapat, Ketan Atul Chakraborty, Mrityunjoy Indian Inst Technol Dept Elect & Elect Commun Engn Kharagpur W Bengal India

ISBN: (数字)9781665482509

ISBN: (纸本)9781665482509

In this paper, we propose a new algorithm named Randomized Simultaneous Hard Thresholding pursuit(RSHTP) for the multiple measurements vector (MMV) problem in compressed sensing. In the proposed algorithm, the gradient is calculated only with respect to few of the signals at each iteration that are chosen randomly. This reduces the computational cost which is significant when the problem size is large. A deterministic convergence analysis is carried out where we present theoretical guarantees using the restricted isometric property (RIP). Simulation studies show that the proposed algorithm enjoys at par performance even at a moderate rate of column selection in each iteration.

关键词： Conferences Computational modeling Signal processing algorithms Signal processing Computational efficiency pursuit algorithms Compressed sensing

来源：评论

学校读者我要写书评

暂无评论

AUTONOMOUS DECENTRALIZED COORDINATION CONTROL FOR FRACTIONATED SPACECRAFT FORMATION RECONFIGURATION BASED ON CYCLIC pursuit STRATEGY

AUTONOMOUS DECENTRALIZED COORDINATION CONTROL FOR FRACTIONAT...

引用

1st International-Academy-of-Astronautics Conference on Dynamics and Control of Space Systems (DyCoSS)

作者： Hu Min Zeng Guoqiang Acad Equipment Beijing Peoples R China

ISBN: (纸本)9780877035879

The current paper investigates the decentralized coordination control problem of fractionated spacecraft and proposes a new concept of finite-time cyclic pursuit algorithm for its practical implementation. Firstly, the relative translational dynamics of fractionated spacecraft and the general formation configuration description method are proposed, and the algebraic graph theory is introduced. Then, three-dimensional finite-time cyclic pursuit algorithms are put forward for both single-integrator kinematics and double-integrator dynamics, which guarantee the convergence of tracking errors in finite time rather than in the asymptotic sense. Finally, numerical simulations on finite-time cyclic pursuit algorithms and formation reconfigurations based on cyclic pursuit strategies are carried out. The simulation results have demonstrated the superior performance of finite-time cyclic pursuit algorithms, including the global fast convergence of tracking errors in finite time and the higher tracking accuracy. Moreover, the cyclic pursuit control is a very promising technique for fractionated spacecraft free flying, because it is leaderless, easy for implementation, and most important of all, it only requires the local information for control.

关键词： pursuit algorithms Spacecraft Decentralized Graph theory Standing Orders formation local information. follow error cyclic Information control

来源：评论

学校读者我要写书评

暂无评论

BAYESIAN pursuit ALGORITHM FOR SPARSE REPRESENTATION

BAYESIAN PURSUIT ALGORITHM FOR SPARSE REPRESENTATION

引用

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Zayyani, H. Babaie-Zadeh, M. Jutten, C. Sharif Univ Technol Adv Commun Res Inst Dept Elect Engn Tehran Iran GIPSA Lab Grenoble France

ISBN: (纸本)9781424423538

In this paper, we propose a Bayesian pursuit algorithm for sparse representation. It uses both the simplicity of the pursuit algorithms and optimal Bayesian framework to determine active atoms in sparse representation of a signal. We show that using Bayesian Hypothesis testing to determine the active atoms from the correlations leads to an efficient activity measure. Simulation results show that our suggested algorithm has better performance among the algorithms which have been implemented in our simulations in most of the cases.

关键词： Sparse representation Sparse Component Analysis (SCA) Compressed Sensing (CS) pursuit algorithms

来源：评论

学校读者我要写书评

暂无评论

Topologically distinct trajectory predictions for probabilistic pursuit

Topologically distinct trajectory predictions for probabilis...

引用

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) / Workshop on Machine Learning Methods for High-Level Cognitive Capabilities in Robotics

作者： Shkurti, Florian Dudek, Gregory McGill Univ Sch Comp Sci Montreal PQ Canada McGill Univ Ctr Intelligent Machines Montreal PQ Canada

ISBN: (纸本)9781538626825

We address the integrated planning and control problem that enables a single follower robot (the "photographer") to maintain a moving target (the "subject") in its field of view for as long as possible. We propose a real-time pursuit algorithm that seamlessly handles the often neglected, yet unavoidable, scenario in which the target escapes the follower's field of view;a scenario that simple, reactive controllers are ill-equipped to handle. Our algorithm aims to minimize the expected time until visual contact is re-established, which enables the photographer to track the subject for as long as possible, even in the presence of loss of visibility. At the core of our pursuit algorithm is an efficient method for sampling plausible trajectories from different homotopy classes. We do this by generating topologically distinct shortest paths by using the Voronoi diagram. We use these paths to make informed, model-based predictions of the likely future locations of the target, given a history of observations. Given these predictions, our algorithm produces pursuit trajectories that approximately minimize the expected time to recover visual contact. We show that constraining the predictive pursuit problem to the space of homotopy classes condenses the expanse of possibilities that our algorithm must consider, which enables target tracking in large occupancy grids, as opposed to many POMDP methods that are constrained to small environments. We benchmark the tracking behavior of our algorithm against the baseline of human subjects who performed the same set of pursuit tasks in simulation, as well as against two other pursuit algorithms that only take into account paths from a single homotopy class. We show that considering homotopy alternatives in 2D pursuit improves the tracking performance and that our algorithm does at least as well as humans in most pursuit scenarios.

关键词： Visualization Robots Trajectory Planning Target tracking Games pursuit algorithms

来源：评论

学校读者我要写书评

暂无评论

A Novel Matching pursuit Algorithm with Adaptive Subdictionary

A Novel Matching Pursuit Algorithm with Adaptive Subdictiona...

引用

9th International Conference on Signal Processing

作者： Guo Jinku Liu Guangbin Yang Xiaojun Xian Res Inst Hitech Dept Automat Xian 710025 Peoples R China Tsinghua Univ Dept Automat Beijing 100084 Peoples R China

ISBN: (纸本)9781424421787

Sparse decomposition of a signal can be obtained by decomposing the signal in an overcomplete dictionary. Matching pursuit (MP) is a greedy algorithm to realize such decompositions, and has been widely studied in many signal processing researches. Through exploiting the structure shared by these so-called adaptive decomposition dictionaries, we propose a novel MP algorithm which can reduce computation and storage requirement dramatically.

关键词： Matching pursuit algorithms pursuit algorithms Dictionaries Signal processing algorithms Time frequency analysis Automation Signal resolution Cities and towns Greedy algorithms Adaptive signal processing

来源：评论

学校读者我要写书评

暂无评论

Instantaneous frequency estimation via conditional spectral moments and matching pursuit decomposition 3

Instantaneous frequency estimation via conditional spectral ...

引用

3rd IEEE International Symposium on Signal Processing and Information Technology

作者： Ghofrani, S McLernon, DC Ayatollahi, A Univ Leeds Sch Elect & Elect Engn Leeds LS2 9JT W Yorkshire England

ISBN: (纸本)0780382927

The conditional spectral moments are important concepts in signal analysis. In this paper, we decompose a non-stationary signal by use of the matching pursuit algorithm, through using two different types of dictionaries (i.e., the Gaussian and damped sinusoids dictionaries). Then we give expressions for the first and second conditional spectral moments, which are generalizations of the ideas of instantaneous frequency and instantaneous bandwidth. Although in many cases the second conditional moment is not positive and this makes the usual interpretation of this quantity impossible, in this paper we will prove that with matching pursuit decomposition, by using the Gaussian or damped sinusoids dictionaries, the second conditional moment is always positive. In addition, we show that the first moment closely estimates the true instantaneous frequency of the signal.

关键词： Bandwidth Dictionaries Energy resolution Frequency estimation Matching pursuit algorithms pursuit algorithms Signal analysis Signal mapping Signal resolution Time frequency analysis

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：