• Transactions on Control, Automation and Systems Engineering
• /
• v.1 no.1
• /
• pp.8-15
• /
• 1999

One of the main reason advocating redundant manipulators' superiority in application is that they can afford to optimize a dexterity measure, for example the manipulability measure. However, to obtain the gradient of the manipulability is not an easy task in case of general manipulator with high degrees of redundancy. This article proposes a method to compute the gradient of the manipulability, based on recursive algorithm to compute the Jacobian and its derivative using Denavit-Hartenberg parameters only. To characterize the null motion of redundant manipulators, the null space matrix using square minors of the Jacobian is also proposed. With these capabilities, the inverse kinematics of a redundant manipulator system can be done automatically. The result is easily extended to dual manipulator system using the relative kinematics.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.5
• /
• pp.3132-3138
• /
• 2014

In many communication systems, intersymbol interference, DC and impulsive noise are hard-to-solve problems. For the purpose of cancelling such interferences, the concept of lagged cross-correlation of probability has been used for blind equalization. However, this algorithm has a large burden of computation. In this paper, a recursive method of the algorithm based on the lagged probability correlation is proposed. The summation operation in the calculation of gradient of the cost is transformed into a recursive gradient calculation. The recursive method shows to reduce the high computational complexity of the algorithm from O(NM) to O(M) for M symbols and N block data having advantages in implementation while keeping the robustness against those interferences. From the results of the simulation, the proposed method yields the same learning performance with reduced computation complexity.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.13 no.3
• /
• pp.186-199
• /
• 2013

We pose pattern classification as a density estimation problem where we consider mixtures of generative models under partially labeled data setups. Unlike traditional approaches that estimate density everywhere in data space, we focus on the density along the decision boundary that can yield more discriminative models with superior classification performance. We extend our earlier work on the recursive estimation method for discriminative mixture models to semi-supervised learning setups where some of the data points lack class labels. Our model exploits the mixture structure in the functional gradient framework: it searches for the base mixture component model in a greedy fashion, maximizing the conditional class likelihoods for the labeled data and at the same time minimizing the uncertainty of class label prediction for unlabeled data points. The objective can be effectively imposed as individual mixture component learning on weighted data, hence our mixture learning typically becomes highly efficient for popular base generative models like Gaussians or hidden Markov models. Moreover, apart from the expectation-maximization algorithm, the proposed recursive estimation has several advantages including the lack of need for a pre-determined mixture order and robustness to the choice of initial parameters. We demonstrate the benefits of the proposed approach on a comprehensive set of evaluations consisting of diverse time-series classification problems in semi-supervised scenarios.

• The Journal of the Acoustical Society of Korea
• /
• v.38 no.3
• /
• pp.283-289
• /
• 2019

RLS (Recursive-least-squares) algorithm is known to have good convergence and excellent error level after convergence. However, there is a disadvantage that numerical instability is included in the algorithm due to inverse matrix calculation. In this paper, we propose an algorithm with no matrix inversion to avoid the instability aforementioned. The proposed algorithm still keeps the same convergence performance. In the proposed algorithm, we adopt an averaged gradient-based step size as a self-adjusted step size. In addition, a variable forgetting factor is introduced to provide superior performance for time-varying channel estimation. Through simulations, we compare performance with conventional RLS and show its equivalency. It also shows the merit of the variable forgetting factor in time-varying channels.

• Journal of Institute of Control, Robotics and Systems
• /
• v.7 no.12
• /
• pp.985-992
• /
• 2001

This paper proposes a self-organizing fuzzy modeling(SOFUM)which an create a new hyperplane shaped cluster and adjust parameters of the fuzzy model in repetition. The suggested algorithm SOFUM is composed of four steps: coarse tuning. fine tuning cluster creation and optimization of learning rates. In the coarse tuning fuzzy C-regression model(FCRM) clustering and weighted recursive least squared (WRLS) algorithm are used and in the fine tuning gradient descent algorithm is used to adjust parameters of the fuzzy model precisely. In the cluster creation, a new hyperplane shaped cluster is created by applying multiple regression to input/output data with relatively large fuzzy entropy based on parameter tunings of fuzzy model. And learning rates are optimized by utilizing meiosis-genetic algorithm in the optimization of learning rates To check the effectiveness of the suggested algorithm two examples are examined and the performance of the identified fuzzy model is demonstrated via computer simulation.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.40 no.2
• /
• pp.247-252
• /
• 2015

Adaptive algorithms based on the criterion of zero-error probability (ZEP) have robustness to impulsive noise and their decision feedback (DF) versions are known to compensate effectively for severe multipath channel distortions. However the ZEP-DF algorithm computes several summation operations at each iteration time for each filter section and this plays an obstacle role in practical implementation. In this paper, the ZEP-DF with recursive gradient estimation (RGE) method is proposed and shown to reduce the computational burden of O(N) to a constant which is independent of the sample size N. Also the weight update of the initial state and the steady state is a continuous process without bringing about any propagation of gradient estimation error in DF structure.

• Transactions on Control, Automation and Systems Engineering
• /
• v.4 no.2
• /
• pp.124-129
• /
• 2002

This paper demonstrates that the largest Lyapunov exponent λ of recurrent neural networks can be controlled efficiently by a stochastic gradient method. An essential core of the proposed method is a novel stochastic approximate formulation of the Lyapunov exponent λ as a function of the network parameters such as connection weights and thresholds of neural activation functions. By a gradient method, a direct calculation to minimize a square error (λ - λ$\^$obj/)$^2$, where λ$\^$obj/ is a desired exponent value, needs gradients collection through time which are given by a recursive calculation from past to present values. The collection is computationally expensive and causes unstable control of the exponent for networks with chaotic dynamics because of chaotic instability. The stochastic formulation derived in this paper gives us an approximation of the gradients collection in a fashion without the recursive calculation. This approximation can realize not only a faster calculation of the gradient, but also stable control for chaotic dynamics. Due to the non-recursive calculation. without respect to the time evolutions, the running times of this approximation grow only about as N$^2$ compared to as N$\^$5/T that is of the direct calculation method. It is also shown by simulation studies that the approximation is a robust formulation for the network size and that proposed method can control the chaos dynamics in recurrent neural networks efficiently.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.3
• /
• pp.2172-2177
• /
• 2015

The DF-MZEP-CME (decision feedback - maximum zero-error probability for constant modulus errors) algorithm that makes the probability for constant modulus error (CME) close to zero and employs decision feedback (DF) structures shows more improved performance in channel distortion compensation. However the DF-MZEP-CME algorithm has a computational complexity proportional to a sample size for probability estimation and this property plays a role of an obstacle in practical implementation. In this paper, the gradient of DF-MZEP-CME is proposed to be estimated recursively and shown to solve the computational problem by making the algorithm independent of the sample size. For a sample size N, the conventional method has 10N multiplications but the proposed has only 20 regardless of N. Also the recursive gradient estimation for weight update is kept in continuity from the initial state to the steady state without any error propagation.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.5
• /
• pp.3434-3439
• /
• 2015

As a criterion of information theoretic learning, the Euclidean distance (ED) of two error probability distribution functions (minimum ED of error, MEDE) has been adopted in nonlinear (decision feedback, DF) supervised equalizer algorithms and has shown significantly improved performance in severe channel distortion and impulsive noise environments. However, the MEDE-DF algorithm has the problem of heavy computational complexity. In this paper, the recursive ED for MEDE-DF algorithm is derived first, and then the feed-forward and feedback section gradients for weight update are estimated recursively. To prove the effectiveness of the recursive gradient estimation for the MEDE-DF algorithm, the number of multiplications are compared and MSE performance in impulsive noise and underwater communication environments is compared through computer simulation. The ratio of the number of multiplications between the proposed DF and the conventional MEDE-DF algorithm is revealed to be $2(9N+4):2(3N^2+3N)$ for the sample size N with the same MSE learning performance in the impulsive noise and underwater channel environment.

• Journal of Broadcast Engineering
• /
• v.24 no.4
• /
• pp.633-642
• /
• 2019

Recently, the convolution neural network (CNN) model at a single image super-resolution (SISR) have been very successful. The residual learning method can improve training stability and network performance in CNN. In this paper, we propose a SISR using recursive residual network architecture by introducing dense skip connections for learning nonlinear mapping from low-resolution input image to high-resolution target image. The proposed SISR method adopts a method of the recursive residual learning to mitigate the difficulty of the deep network training and remove unnecessary modules for easier to optimize in CNN layers because of the concise and compact recursive network via dense skip connection method. The proposed method not only alleviates the vanishing-gradient problem of a very deep network, but also get the outstanding performance with low complexity of neural network, which allows the neural network to perform training, thereby exhibiting improved performance of SISR method.