基于奇异值分解的广义系统迭代学习控制算法

在对广义系统进行奇异值分解的基础上,研究了一类广义系统的迭代学习控制问题。针对快子系统和慢子系统的特点,分别利用状态误差代入输出误差,得到了一类新的广义系统迭代学习控制算法结构,这一算法是全新的。然后从理论上对所提出的算法进行完整的收敛性分析。分析结果表明,满足给定的收敛条件,系统输出可以渐近地跟踪给定的期望轨迹。     

基于累积量极值的非线性系统盲反卷积

提出一种新的基于累积量极值的非线性系统盲反卷积算法.通过引入中间变量作为隐含观测量,反卷积系统估计转化为两个子系统估计问题.第一个子系统估计是一具有完备指导信号训练集的后非线性系统(post nonlinear system)辨识,第二个子系统的估计则是一般基于累积量极值的线性系统盲反卷积.与其它算法比较,新算法具有明确的收敛性质和快速的收敛速度.     

基于改进BP神经网络的数字识别

针对BP(Back Propagation)神经网络易陷入局部极小、收敛速度慢的缺点,提出了一种新的BP神经网络改进算法.与标准BP算法比较,该系统通过结合附加动量法和自适应学习速率形成新的BP改进算法.附加动量法虽然可以使BP算法避免陷入局部极小,但是对初始值的选取比较敏感,而且选取合适的学习速率比较困难.而自适应学...     

Convergence performance analysis of an adaptive kernel width MCC algorithm

A fixed kernel width in MCC algorithm imposes a trade-off among robustness, convergence rate and steady-state accuracy. With a variable kernel width, the adaptive kernel width MCC (AMCC) algorithm can improve the learning speed of the MCC algorithm especially when the initial weight vector is far away from the optimal weight vector. In this paper, the steady-state excess mean square error (EMSE) of the AMCC algorithm is studied based on energy conservation relation. In addition, a novel convergence measure called initial convergence rate is introduced to evaluate the convergence speed at the beginning of the learning. Simulation experiments are carried out to verify the theoretical analysis and confirm the desirable performance of the AMCC algorithm in several different non-Gaussian noise environments.     

Wavelet transform based adaptive filters: analysis and new results

In this paper the wavelet transform is used in an adaptive filtering structure. The coefficients of the adaptive filter are updated by the help of the least mean square (LMS) algorithm. First, the wavelet transform based adaptive filter (WTAF) is described and it is analyzed for its Wiener optimal solution. Then the performance of the WTAF is studied by the help of learning curves for three different convergence factors: (1) constant convergence factor, (2) time-varying convergence factor, and (3) exponentially weighted convergence factor. The exponentially weighted convergence factor is proposed to introduce scale-based variation to the weight update equation. It is shown for two different sets of data that the rate of convergence increases significantly for all three WTAF structures as compared to that of time-domain LMS. The high convergence rates of the WTAF give us reason to expect that it will perform well in tracking rapid changes in a signal     

A dynamic learning neural network for remote sensing applications

The neural network learning process is to adjust the network weights to adapt the selected training data. Based on the polynomial basis function (PBF) modeled neural network that is a modified multilayer perceptrons (MLP) network, a dynamic learning algorithm (DL) is proposed. The presented learning algorithm makes use of the Kalman filtering technique to update the network weights, in the sense that the stochastic characteristics of incoming data sets are implicitly incorporated into the network. The Kalman gains which represent the learning rates of the network weights updating are calculated by using the U-D factorization. By concatenating all of the network weights at each layer to form a long vector such that it can be updated without propagating back, the proposed algorithm improves the performance of convergence to which the backpropagation (BP) learning algorithm often suffers. Numerical illustrations are carried out using two categories of problems: multispectral imagery classification and surface parameters inversion. Results indicates the use of Kalman filtering algorithm not only substantially increases the convergence rate in the learning stage, but also enhances the separability for highly nonlinear boundaries problems, as compared to BP algorithm, suggesting that the proposed DL neural network provides a practical and potential tool for remote sensing applications     

一种新的基于值函数迁移的快速Sarsa算法

知识迁移是当前机器学习领域的一个新的研究热点.其基本思想是通过将经验知识从历史任务到目标任务的迁移,达到提高算法收敛速度和收敛精度的目的.针对当前强化学习领域中经典算法收敛速度慢的问题,提出在学习过程中通过迁移值函数信息,减少算法收敛所需要的样本数量,加快算法的收敛速度.基于强化学习中经典的在策略Sarsa算法的学习框架,结合值函数迁移方法,优化算法初始值函数的设置,提出一种新的基于值函数迁移的快速Sarsa算法--VFT-Sarsa.该算法在执行前期,通过引入自模拟度量方法,在状态空间以及动作空间一致的情况下,对目标任务中的状态与历史任务中的状态之间的距离进行度量,对其中相似并满足一定条件的状态进行值函数迁移,而后再通过学习算法进行学习.将VTF-Sarsa算法用于Random Walk问题,并与经典的Sarsa算法、Q学习算法以及具有较好收敛速度的QV算法进行比较,实验结果表明,该算法在保证收敛精度的基础上,具有更快的收敛速度.     

Detection of multiple sclerosis with visual evoked potentials - anunsupervised computational intelligence system

This paper describes the application of a novel unsupervised pattern recognition system to the classification of the visual evoked potentials (VEPs) of normal and multiple sclerosis (MS) patients. The method combines a traditional statistical feature extractor with a fuzzy clustering method, all implemented in a parallel neural network architecture. The optimization routine, ALOPEX, is used to train the network while decreasing the livelihood of local solutions. The unsupervised system includes a feature extraction and clustering module, trained by the optimization routine ALOPEX. Through maximization of the output variance of each node, and an architecture which excludes redundancy, the feature extraction network retains the most significant Karhunen-Loeve expansion vectors. The clustering module uses a modification to the fuzzy c-means (FCM) clustering algorithms, where ALOPEX adjusts a set of cluster centers to minimize an objective error function. The result combines the power of the FCM algorithms with the advantage of a more global solution from ALOPEX. The new pattern recognition system is used to cluster the VEPs of 13 normal and 12 MS subjects. The classification with this technique can, without supervision, separate the patient population into two groups which largely correspond to the MS and control subject groups. A suitable threshold can be chosen so that the recognizer chooses no false negatives. The use of multiple stimulation patterns appears to improve the reliability of the decision. The reasoning of most neural networks in their decision making cannot easily be extracted upon the completion of training. However, due to the linearity of the network nodes, the cluster prototypes of this unsupervised system can be reconstructed to illustrate the reasoning of the system. In this application, this analysis hints at the usefulness of previously unused portions of the VEP in detecting MS. It also indicates a possible use of the system as a training aide     

A Modified Oja–Xu MCA Learning Algorithm and Its Convergence Analysis

The original Oja-Xu minor component analysis (MCA) learning algorithm is not convergent. This brief shows that by modifying Oja-Xu MCA learning algorithm with a normalization step the modified one could be convergent subject to some conditions satisfied. The convergence of the modified MCA learning algorithm is studied by analyzing the convergence of an associated deterministic discrete time system. Necessary and sufficient conditions for convergence are obtained. Simulations further confirm the results     

An Improved Grasshopper Optimization Algorithm for Global Optimization

We proposes an improved grasshopper algorithm for global optimization problems. Grasshopper optimization algorithm (GOA) is a recently proposed meta-heuristic algorithm inspired by the swarming behav-ior of grasshoppers. The original GOA has some drawbacks, such as slow convergence speed, easily falling into local optimum, and so on. To overcome these shortcomings, we proposes a grasshopper optimization algorithm based on a logistic Chaos maps opposition-based learning strategy and cloud model inertia weight (CCGOA). CCGOA is divided into three stages. The chaos opposition learning initialization strategy is used to initialize the population, so that the population can be evenly distributed in the feasible solution space as much as possible, so as to improve the uniformity and diversity of the initial population distribution of the grasshopper algorithm. The inertia weight cloud model is introduced into the grasshopper algorithm, and different inertia weight strategies are used to adjust the convergence speed of the algorithm. Based on the principle of chaotic logistic maps, local depth search is carried out to reduce the probability of falling into local optimum. Fourteen benchmark functions and an engineering example are used for simulation verification. Experimental results show that the proposed CCGOA algorithm has superior performance in determining the optimal solution of the test function problem.     

Power Line Communication with Network Transmission Data Loss Based on Learning Control

This paper proposed power line communication with transmission data. An iterative learning control method for the power line communication is studied by P-type learning control law. The data packet loss described as a stochastic Bernoulli process. The sufficient conditions are given for the convergence of the proposed algorithm by using the compression mapping method and norm theory. The convergence analysis guarantee the convergence of the tracking error in the sense of the \(\uplambda\)-norm. Finally, numerical simulations illustrate to verify the effectiveness of the proposed learning algorithm.     

一种基于反向学习的约束差分进化算法

差分进化算法是一种结构简单、易用且鲁棒性强的全局搜索启发式优化算法,它可以结合约束处理技术来解决约束优化问题.机器学习在进化算法中,经常可以引导种群的进化,而且被广泛地应用于无约束的差分进化算法中,但对于约束差分进化算法却很少有应用.针对这一情况,提出了一种基于反向学习的约束差分进化算法框架.该算法框架采用基于反向学习的机器学习方法,提高约束差分进化算法的多样性和加速全局收敛速度.最后把该算法框架植入了两个著名的约束差分进化算法:(μ+λ)-CDE和ECHT,并采用CEC 2010的18个Benchmark函数进行了实验评估,实验结果表明:与(μ+λ)-CDE和ECHT相比,植入后的算法具有更强的全局搜索能力、更快的收敛速度和更高的收敛精度.     

基于优先级扫描Dyna结构的贝叶斯Q学习方法

贝叶斯Q学习方法使用概率分布来描述Q值的不确定性,并结合Q值分布来选择动作,以达到探索与利用的平衡。然而贝叶斯Q学习存在着收敛速度慢且收敛精度低的问题。针对上述问题,提出一种基于优先级扫描Dyna结构的贝叶斯Q学习方法—Dyna-PS-BayesQL。该方法主要分为2部分：在学习部分,对环境的状态迁移函数及奖赏函数建模,并使用贝叶斯Q学习更新动作值函数的参数;在规划部分,基于建立的模型,使用优先级扫描方法和动态规划方法对动作值函数进行规划更新,以提高对历史经验信息的利用,从而提升方法收敛速度及收敛精度。将Dyna-PS-BayesQL应用于链问题和迷宫导航问题,实验结果表明,该方法能较好地平衡探索与利用,且具有较优的收敛速度及收敛精度。     

Hybrid social cognitive optimization algorithm for constrained nonlinear programming

To improve the global convergence speed of social cognitive optimization (SCO) algorithm,a hybrid social cognitive optimization (HSCO) algorithm based on elitist strategy and chaotic optimization is pr...     

Nonlinear blind signal separation with intelligent controlled learning

The paper proposes a new nonlinear blind source separation algorithm with hybridisation of fuzzy logic based learning rate control and simulated annealing to improve the global solution search. Benefits of fuzzy systems and simulated annealing are incorporated into a multilayer perceptron network. Fuzzy logic control allows adjustments of learning rate to enhance the rate of convergence of the algorithm. Simulated annealing is implemented to avoid the algorithm becoming trapped in local minima. A simple and computationally efficient method for controlling learning rate and ensuring a global solution is proposed. The performance of the proposed algorithm in terms of convergence of entropy, is studied alongside other techniques of learning rate adaptation. Simulations show that the proposed nonlinear algorithm outperforms other existing nonlinear algorithms based on fixed learning rates.     

BP网络不平衡训练样本集的有效学习算法

分析了BP网络标准反传学习算法对不平衡样本集训练速度慢的原因,研究了如何改进其学习算法来加速训练速度,并通过实验对上述理论进行验证。     

峭度自适应学习率的盲信源分离

本文提出了一种自适应学习率盲信源分离的自然梯度算法,自适应学习率仅依赖于神经网络输出峭度平方和的负指数.开始阶段由于小的峭度,学习率大收敛速度快.之后,随着峭度变大,学习率慢慢变小,产生小的稳态误差.在线性无记忆混合的情况下,用欠高斯信源进行的模拟实验表明,与固定学习率相比,本文提出的峭度自适应学习率盲信源分离算法具有收敛速度快和稳态误差小的特点.     

Multigradient: a new neural network learning algorithm for patternclassification

The authors propose a new learning algorithm for multilayer feedforward neural networks, which converges faster and achieves a better classification accuracy than the conventional backpropagation learning algorithm for pattern classification. In the conventional backpropagation learning algorithm, weights are adjusted to reduce the error or cost function that reflects the differences between the computed and the desired outputs. In the proposed learning algorithm, the authors view each term of the output layer as a function of weights and adjust the weights directly so that the output neurons produce the desired outputs. Experiments with remotely sensed data show the proposed algorithm consistently performs better than the conventional backpropagation learning algorithm in terms of classification accuracy and convergence speed     

基于代理函数优化的稀疏性字典学习

稀疏性字典学习是指对在某个已知的基字典上具有稀疏表示的字典的学习.论文利用块松弛思想,将稀疏性字典学习问题转化为字典和系数的分别优化问题,利用代理函数优化方法分别对固定字典和固定系数情况下的目标函数进行优化处理,得到固定字典情况下的系数更新算法和固定系数情况下的字典更新算法,进而得到稀疏性字典学习算法.理论分析说明了本...     

同伦方法在图像稀疏去噪中的应用

本文在深入研究稀疏表示和字典学习理论的基础上,建立了图像去噪模型并提出一种新的图像去噪算法。该算法采用同伦方法学习字典,充分利用了同伦方法收敛速度快以及对信号的恢复准确度高的特点。之后利用 OMP 算法求出带噪图像在该字典下的稀疏表示系数,并结合稀疏去噪模型实现对图像的去噪。实验结果显示本文算法在不同的噪声环境下具有较好的去噪效果,同时在与 K-SVD 算法关于收敛速度比较的实验中,实验结果充分显示了使用同伦算法学习字典在收敛速度上的优势。