Search results for: Affine Invariant

Authors: Bingquan Shen

Abstract:

Convolutional Neural Networks (CNN) have demonstrated their effectiveness in synthesizing 3D views of object instances at various viewpoints. Given the problem where one have limited viewpoints of a particular object for classification, we present a pose normalization architecture to transform the object to existing viewpoints in the training dataset before classification to yield better classification performance. We have demonstrated that this Pose Normalization Network (PNN) can capture the style of the target object and is able to re-render it to a desired viewpoint. Moreover, we have shown that the PNN improves the classification result for the 3D chairs dataset and ShapeNet airplanes dataset when given only images at limited viewpoint, as compared to a CNN baseline.

Keywords: Convolutional neural networks, object classification, pose normalization, viewpoint invariant.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1120

Authors: Adnan A. Y. Mustafa

Abstract:

In this paper we present a quick technique to measure the similarity between binary images. The technique is based on a probabilistic mapping approach and is fast because only a minute percentage of the image pixels need to be compared to measure the similarity, and not the whole image. We exploit the power of the Probabilistic Matching Model for Binary Images (PMMBI) to arrive at an estimate of the similarity. We show that the estimate is a good approximation of the actual value, and the quality of the estimate can be improved further with increased image mappings. Furthermore, the technique is image size invariant; the similarity between big images can be measured as fast as that for small images. Examples of trials conducted on real images are presented.

Keywords: Big images, binary images, similarity, matching.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 919

Authors: Vesna Kirandziska, Nevena Ackovska, Ana Madevska Bogdanova

Abstract:

The problem of emotion recognition is a challenging problem. It is still an open problem from the aspect of both intelligent systems and psychology. In this paper, both voice features and facial features are used for building an emotion recognition system. A Support Vector Machine classifiers are built by using raw data from video recordings. In this paper, the results obtained for the emotion recognition are given, and a discussion about the validity and the expressiveness of different emotions is presented. A comparison between the classifiers build from facial data only, voice data only and from the combination of both data is made here. The need for a better combination of the information from facial expression and voice data is argued.

Keywords: Emotion recognition, facial recognition, signal processing, machine learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2018

Authors: Ehab F. Badran, Guoxiang Gu

Abstract:

We consider optimal channel equalization for MIMO (multi-input/multi-output) time-varying channels in the sense of MMSE (minimum mean-squared-error), where the observation noise can be non-stationary. We show that all ZF (zero-forcing) receivers can be parameterized in an affine form which eliminates completely the ISI (inter-symbol-interference), and optimal channel equalizers can be designed through minimization of the MSE (mean-squarederror) between the detected signals and the transmitted signals, among all ZF receivers. We demonstrate that the optimal channel equalizer is a modified Kalman filter, and show that under the AWGN (additive white Gaussian noise) assumption, the proposed optimal channel equalizer minimizes the BER (bit error rate) among all possible ZF receivers. Our results are applicable to optimal channel equalization for DWMT (discrete wavelet multitone), multirate transmultiplexers, OFDM (orthogonal frequency division multiplexing), and DS (direct sequence) CDMA (code division multiple access) wireless data communication systems. A design algorithm for optimal channel equalization is developed, and several simulation examples are worked out to illustrate the proposed design algorithm.

Keywords: Channel equalization, Kalman filtering, Time-varying systems.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1834

Authors: Shao-Tzu Huang, Chen-Chien Hsu, Wei-Yen Wang

Abstract:

Matching high dimensional features between images is computationally expensive for exhaustive search approaches in computer vision. Although the dimension of the feature can be degraded by simplifying the prior knowledge of homography, matching accuracy may degrade as a tradeoff. In this paper, we present a feature matching method based on k-means algorithm that reduces the matching cost and matches the features between images instead of using a simplified geometric assumption. Experimental results show that the proposed method outperforms the previous linear exhaustive search approaches in terms of the inlier ratio of matched pairs.

Keywords: Feature matching, k-means clustering, scale invariant feature transform, linear exhaustive search.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1086

Authors: Mahmood M. Al-Imam, M. Mustafa

Abstract:

In this paper, we present an experimental testing for a new algorithm that determines an optimal controller-s coefficients for output variance reduction related to Linear Time Invariant (LTI) Systems. The algorithm features simplicity in calculation, generalization to minimal and non-minimal phase systems, and could be configured to achieve reference tracking as well as variance reduction after compromising with the output variance. An experiment of DCmotor velocity control demonstrates the application of this new algorithm in designing the controller. The results show that the controller achieves minimum variance and reference tracking for a preset velocity reference relying on an identified model of the motor.

Keywords: Output variance, minimum variance, overparameterization, DC-Motor.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1359

Authors: B. M. Mohan, Sanjeeb Kumar Kar

Abstract:

In this paper a unified approach via block-pulse functions (BPFs) or shifted Legendre polynomials (SLPs) is presented to solve the linear-quadratic-Gaussian (LQG) control problem. Also a recursive algorithm is proposed to solve the above problem via BPFs. By using the elegant operational properties of orthogonal functions (BPFs or SLPs) these computationally attractive algorithms are developed. To demonstrate the validity of the proposed approaches a numerical example is included.

Keywords: Linear quadratic Gaussian control, linear quadratic estimator, linear quadratic regulator, time-invariant systems, orthogonal functions, block-pulse functions, shifted legendre polynomials.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1859

Authors: Yasuo Obikane

Abstract:

This work is to study a roll of the fluctuating density gradient in the compressible flows for the computational fluid dynamics (CFD). A new anisotropy tensor with the fluctuating density gradient is introduced, and is used for an invariant modeling technique to model the turbulent density gradient correlation equation derived from the continuity equation. The modeling equation is decomposed into three groups: group proportional to the mean velocity, and that proportional to the mean strain rate, and that proportional to the mean density. The characteristics of the correlation in a wake are extracted from the results by the two dimensional direct simulation, and shows the strong correlation with the vorticity in the wake near the body. Thus, it can be concluded that the correlation of the density gradient is a significant parameter to describe the quick generation of the turbulent property in the compressible flows.

Keywords: Turbulence Modeling , Density Gradient Correlation, Compressible

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1446

Authors: Kittiwann Nimkerdphol, Masahiro Nakagawa

Abstract:

Biological reactions of individuals of a testing animal to toxic substance are unique and can be used as an indication of the existing of toxic substance. However, to distinguish such phenomenon need a very complicate system and even more complicate to analyze data in 3 dimensional. In this paper, a system to evaluate in vitro biological activities to acute toxicity of stochastic self-affine non-stationary signal of 3D goldfish swimming by using fractal analysis is introduced. Regular digital camcorders are utilized by proposed algorithm 3DCCPC to effectively capture and construct 3D movements of the fish. A Critical Exponent Method (CEM) has been adopted as a fractal estimator. The hypothesis was that the swimming of goldfish to acute toxic would show the fractal property which related to the toxic concentration. The experimental results supported the hypothesis by showing that the swimming of goldfish under the different toxic concentration has fractal properties. It also shows that the fractal dimension of the swimming related to the pH value of FD Ôëê 0.26pH + 0.05. With the proposed system, the fish is allowed to swim freely in all direction to react to the toxic. In addition, the trajectories are precisely evaluated by fractal analysis with critical exponent method and hence the results exhibit with much higher degree of confidence.

Keywords: 3D locomotion, bioassay, critical exponent method, CEM, fractal analysis, goldfish.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1724

Authors: Integral square error, Least-squares, Markovparameters, Moment matching, Order reduction.

Abstract:

The concept of order reduction by least-squares moment matching and generalised least-squares methods has been extended about a general point ?a?, to obtain the reduced order models for linear, time-invariant dynamic systems. Some heuristic criteria have been employed for selecting the linear shift point ?a?, based upon the means (arithmetic, harmonic and geometric) of real parts of the poles of high order system. It is shown that the resultant model depends critically on the choice of linear shift point ?a?. The validity of the criteria is illustrated by solving a numerical example and the results are compared with the other existing techniques.

Keywords: Integral square error, Least-squares, Markovparameters, Moment matching, Order reduction.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1694

Authors: Anuj Abraham, N. Pappa, Daniel Honc, Rahul Sharma

Abstract:

In this paper, an analysis of some model order reduction techniques is presented. A new hybrid algorithm for model order reduction of linear time invariant systems is compared with the conventional techniques namely Balanced Truncation, Hankel Norm reduction and Dominant Pole Algorithm (DPA). The proposed hybrid algorithm is known as Clustering Dominant Pole Algorithm (CDPA), is able to compute the full set of dominant poles and its cluster center efficiently. The dominant poles of a transfer function are specific eigenvalues of the state space matrix of the corresponding dynamical system. The effectiveness of this novel technique is shown through the simulation results.

Keywords: Balanced truncation, Clustering, Dominant pole, Hankel norm, Model reduction.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2687

Authors: Fatima Akhmedova, Simon Liao

Abstract:

One of the most critical decision points in the design of a face recognition system is the choice of an appropriate face representation. Effective feature descriptors are expected to convey sufficient, invariant and non-redundant facial information. In this work we propose a set of Hahn moments as a new approach for feature description. Hahn moments have been widely used in image analysis due to their invariance, nonredundancy and the ability to extract features either globally and locally. To assess the applicability of Hahn moments to Face Recognition we conduct two experiments on the Olivetti Research Laboratory (ORL) database and University of Notre-Dame (UND) X1 biometric collection. Fusion of the global features along with the features from local facial regions are used as an input for the conventional k-NN classifier. The method reaches an accuracy of 93% of correctly recognized subjects for the ORL database and 94% for the UND database.

Keywords: Face Recognition, Hahn moments, Recognition-by-parts, Time-lapse.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1777

Authors: Bibhya Sharma, Jito Vanualailai, Jai Raj

Abstract:

This paper proposes a new obstacle and collision avoidance control laws for a three-dimensional swarm of boids. The swarm exhibit collective emergent behaviors whilst avoiding the obstacles in the workspace. While flocking, animals group up in order to do various tasks and even a greater chance of evading predators. A generalized algorithms for attraction to the centroid, inter-individual swarm avoidance and obstacle avoidance is designed in this paper. We present a set of new continuous time-invariant velocity control laws is presented which is formulated via the Lyapunov-based control scheme. The control laws proposed in this paper also ensures practical stability of the system. The effectiveness of the proposed control laws is demonstrated via computer simulations

 

Keywords: Lyapunov-based Control Scheme, Motion planning, Practical stability, Swarm.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2494

Authors: Say Wei Foo, Qi Dong

Abstract:

In this paper, a novel feature-based image watermarking scheme is proposed. Zernike moments which have invariance properties are adopted in the scheme. In the proposed scheme, feature points are first extracted from host image and several circular patches centered on these points are generated. The patches are used as carriers of watermark information because they can be regenerated to locate watermark embedding positions even when watermarked images are severely distorted. Zernike transform is then applied to the patches to calculate local Zernike moments. Dither modulation is adopted to quantize the magnitudes of the Zernike moments followed by false alarm analysis. Experimental results show that quality degradation of watermarked image is visually transparent. The proposed scheme is very robust against image processing operations and geometric attacks.

Keywords: Image watermarking, Zernike moments, Featurepoint, Invariance, Robustness.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1848

Authors: J. Ruiz-Pinales, J. J. Acosta-Reyes, A. Salazar-Garibay, R. Jaime-Rivas

Abstract:

In this paper, we present a new method for incorporating global shift invariance in support vector machines. Unlike other approaches which incorporate a feature extraction stage, we first scale the image and then classify it by using the modified support vector machines classifier. Shift invariance is achieved by replacing dot products between patterns used by the SVM classifier with the maximum cross-correlation value between them. Unlike the normal approach, in which the patterns are treated as vectors, in our approach the patterns are treated as matrices (or images). Crosscorrelation is computed by using computationally efficient techniques such as the fast Fourier transform. The method has been tested on the ORL face database. The tests indicate that this method can improve the recognition rate of an SVM classifier.

Keywords: Face recognition, support vector machines, shiftinvariance, image registration.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1757

Authors: S. N. Deepa, G. Sugumaran

Abstract:

This paper proposes the novel model order formulation scheme to design a discrete PID controller for higher order linear time invariant discrete systems. Modified PSO (MPSO) based model order formulation technique has used to obtain the successful formulated second order system. PID controller is tuned to meet the desired performance specification by using pole-zero cancellation and proposed design procedures. Proposed PID controller is attached with both higher order system and formulated second order system. System specifications are tabulated and closed loop response is observed for stabilization process. The proposed method is illustrated through numerical examples from literature.

Keywords: Discrete PID controller, Model Order Formulation, Modified Particle Swarm Optimization, Pole-Zero Cancellation

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1613

Authors: Shimal Das, Dibyendu Ghoshal

Abstract:

Fractal based digital image compression is a specific technique in the field of color image. The method is best suited for irregular shape of image like snow bobs, clouds, flame of fire; tree leaves images, depending on the fact that parts of an image often resemble with other parts of the same image. This technique has drawn much attention in recent years because of very high compression ratio that can be achieved. Hybrid scheme incorporating fractal compression and speedup techniques have achieved high compression ratio compared to pure fractal compression. Fractal image compression is a lossy compression method in which selfsimilarity nature of an image is used. This technique provides high compression ratio, less encoding time and fart decoding process. In this paper, fractal compression with quad tree and DCT is proposed to compress the color image. The proposed hybrid schemes require four phases to compress the color image. First: the image is segmented and Discrete Cosine Transform is applied to each block of the segmented image. Second: the block values are scanned in a zigzag manner to prevent zero co-efficient. Third: the resulting image is partitioned as fractals by quadtree approach. Fourth: the image is compressed using Run length encoding technique.

Keywords: Fractal coding, Discrete Cosine Transform, Iterated Function System (IFS), Affine Transformation, Run length encoding.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1570

Authors: S. N. Deepa, G. Sugumaran

Abstract:

This paper proposes a new algebraic scheme to design a PID controller for higher order linear time invariant continuous systems. Modified PSO (MPSO) based model order formulation techniques have applied to obtain the effective formulated second order system. A controller is tuned to meet the desired performance specification by using pole-zero cancellation method. Proposed PID controller is attached with both higher order system and formulated second order system. The closed loop response is observed for stabilization process and compared with general PSO based formulated second order system. The proposed method is illustrated through numerical example from literature.

Keywords: Higher order systems, model order formulation, modified particle swarm optimization, PID controller, pole-zero cancellation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 5028

Authors: Hossein Sahoolizadeh, Mahdi Rahimi, Hamid Dehghani

Abstract:

We introduce an algorithm based on the morphological shared-weight neural network. Being nonlinear and translation-invariant, the MSNN can be used to create better generalization during face recognition. Feature extraction is performed on grayscale images using hit-miss transforms that are independent of gray-level shifts. The output is then learned by interacting with the classification process. The feature extraction and classification networks are trained together, allowing the MSNN to simultaneously learn feature extraction and classification for a face. For evaluation, we test for robustness under variations in gray levels and noise while varying the network-s configuration to optimize recognition efficiency and processing time. Results show that the MSNN performs better for grayscale image pattern classification than ordinary neural networks.

Keywords: Face recognition, Neural Networks, Multi-layer Perceptron, masking.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1516

Authors: Afaf Alharbi, Qianni Zhang

Abstract:

The identification of malignant tissue in histopathological slides holds significant importance in both clinical settings and pathology research. This paper presents a methodology aimed at automatically categorizing cancerous tissue through the utilization of a multiple instance learning framework. This framework is specifically developed to acquire knowledge of the Bernoulli distribution of the bag label probability by employing neural networks. Furthermore, we put forward a neural network-based permutation-invariant aggregation operator, equivalent to attention mechanisms, which is applied to the multi-instance learning network. Through empirical evaluation on an openly available colon cancer histopathology dataset, we provide evidence that our approach surpasses various conventional deep learning methods.

Keywords: Attention Multiple Instance Learning, Multiple Instance Learning, transfer learning, histopathological slides, cancer tissue classification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 221

Authors: Ekprasit Promtun, Sridhar Seshagiri

Abstract:

This paper considers the control of the longitudinal flight dynamics of an F-16 aircraft. The primary design objective is model-following of the pitch rate q, which is the preferred system for aircraft approach and landing. Regulation of the aircraft velocity V (or the Mach-hold autopilot) is also considered, but as a secondary objective. The problem is challenging because the system is nonlinear, and also non-affine in the input. A sliding mode controller is designed for the pitch rate, that exploits the modal decomposition of the linearized dynamics into its short-period and phugoid approximations. The inherent robustness of the SMC design provides a convenient way to design controllers without gain scheduling, with a steady-state response that is comparable to that of a conventional polynomial based gain-scheduled approach with integral control, but with improved transient performance. Integral action is introduced in the sliding mode design using the recently developed technique of “conditional integrators", and it is shown that robust regulation is achieved with asymptotically constant exogenous signals, without degrading the transient response. Through extensive simulation on the nonlinear multiple-input multiple-output (MIMO) longitudinal model of the F-16 aircraft, it is shown that the conditional integrator design outperforms the one based on the conventional linear control, without requiring any scheduling.

Keywords: Sliding-mode Control, Integral Control, Model Following, F-16 Longitudinal Dynamics, Pitch-Rate Control.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3220

Authors: D. R. Ambika, K. R. Radhika, D. Seshachalam

Abstract:

Unconstrained authentication is an important component for personal automated systems and human-computer interfaces. Existing solutions mostly use face as the primary object of analysis. The performance of face-based systems is largely determined by the extent of deformation caused in the facial region and amount of useful information available in occluded face images. Periocular region is a useful portion of face with discriminative ability coupled with resistance to deformation. A reliable portion of periocular area is available for occluded images. The present work demonstrates that joint representation of periocular texture and periocular structure provides an effective expression and poses invariant representation. The proposed methodology provides an effective and compact description of periocular texture and shape. The method is tested over four benchmark datasets exhibiting varied acquisition conditions.

Keywords: Periocular authentication, Zernike moments, LBPV, shape and texture fusion.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 942

Authors: Alima Damak Masmoudi, Randa Boukhris Trabelsi, Dorra Sellami Masmoudi

Abstract:

Single biometric modality recognition is not able to meet the high performance supplies in most cases with its application become more and more broadly. Multimodal biometrics identification represents an emerging trend recently. This paper investigates a novel algorithm based on fusion of both fingerprint and fingervein biometrics. For both biometric recognition, we employ the Monogenic Local Binary Pattern (MonoLBP). This operator integrate the orginal LBP (Local Binary Pattern ) with both other rotation invariant measures: local phase and local surface type. Experimental results confirm that a weighted sum based proposed fusion achieves excellent identification performances opposite unimodal biometric systems. The AUC of proposed approach based on combining the two modalities has very close to unity (0.93).

Keywords: fingerprint, fingervein, LBP, MonoLBP, fusion, biometric trait.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2391

Authors: Ephraim Gower, Thato Tsalaile, Monageng Kgwadi, Malcolm Hawksford.

Abstract:

This paper introduces a new point estimation algorithm, with particular focus on coherent noise suppression, given several measurements of the device under test where it is assumed that 1) the noise is first-order stationery and 2) the device under test is linear and time-invariant. The algorithm exploits the robustness of the Pitman estimator of the Cauchy location parameter through the initial scaling of the test signal by a centred Gaussian variable of predetermined variance. It is illustrated through mathematical derivations and simulation results that the proposed algorithm is more accurate and consistently robust to outliers for different tailed density functions than the conventional methods of sample mean (coherent averaging technique) and sample median search.

Keywords: Central limit theorem, Fisher-Cramer Rao, gamma function, Pitman estimator.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1922

Authors: O. Abusaeeda, J. P. O Evans, D. Downes

Abstract:

We demonstrate the synthesis of intermediary views within a sequence of color encoded, materials discriminating, X-ray images that exhibit animated depth in a visual display. During the image acquisition process, the requirement for a linear X-ray detector array is replaced by synthetic image. Scale Invariant Feature Transform, SIFT, in combination with material segmented morphing is employed to produce synthetic imagery. A quantitative analysis of the feature matching performance of the SIFT is presented along with a comparative study of the synthetic imagery. We show that the total number of matches produced by SIFT reduces as the angular separation between the generating views increases. This effect is accompanied by an increase in the total number of synthetic pixel errors. The trends observed are obtained from 15 different luggage items. This programme of research is in collaboration with the UK Home Office and the US Dept. of Homeland Security.

Keywords: X-ray, kinetic depth, view synthesis, KDE

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1662

Authors: Sudipta Majumdar

Abstract:

This paper presents a method for identification of a linear time invariant (LTI) autonomous all pole system using singular value decomposition. The novelty of this paper is two fold: First, MUSIC algorithm for estimating complex frequencies from real measurements is proposed. Secondly, using the proposed algorithm, we can identify the coefficients of differential equation that determines the LTI system by switching off our input signal. For this purpose, we need only to switch off the input, apply our complex MUSIC algorithm and determine the coefficients as symmetric polynomials in the complex frequencies. This method can be applied to unstable system and has higher resolution as compared to time series solution when, noisy data are used. The classical performance bound, Cramer Rao bound (CRB), has been used as a basis for performance comparison of the proposed method for multiple poles estimation in noisy exponential signal.

Keywords: MUSIC algorithm, Cramer Rao bound, frequency estimation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 934

Authors: Nicholas D. Assimakis

Abstract:

In linear estimation, the traditional Kalman filter uses the Kalman filter gain in order to produce estimation and prediction of the n-dimensional state vector using the m-dimensional measurement vector. The computation of the Kalman filter gain requires the inversion of an m x m matrix in every iteration. In this paper, a variation of the Kalman filter eliminating the Kalman filter gain is proposed. In the time varying case, the elimination of the Kalman filter gain requires the inversion of an n x n matrix and the inversion of an m x m matrix in every iteration. In the time invariant case, the elimination of the Kalman filter gain requires the inversion of an n x n matrix in every iteration. The proposed Kalman filter gain elimination algorithm may be faster than the conventional Kalman filter, depending on the model dimensions.

Keywords: Discrete time, linear estimation, Kalman filter, Kalman filter gain.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 638

Authors: Chenxue Yang, Mao Ye, Zijian Liu, Tao Li, Jiao Bao

Abstract:

Non-negative matrix factorization (NMF) is a useful computational method to find basis information of multivariate nonnegative data. A popular approach to solve the NMF problem is the multiplicative update (MU) algorithm. But, it has some defects. So the columnwisely alternating gradient (cAG) algorithm was proposed. In this paper, we analyze convergence of the cAG algorithm and show advantages over the MU algorithm. The stability of the equilibrium point is used to prove the convergence of the cAG algorithm. A classic model is used to obtain the equilibrium point and the invariant sets are constructed to guarantee the integrity of the stability. Finally, the convergence conditions of the cAG algorithm are obtained, which help reducing the evaluation time and is confirmed in the experiments. By using the same method, the MU algorithm has zero divisor and is convergent at zero has been verified. In addition, the convergence conditions of the MU algorithm at zero are similar to that of the cAG algorithm at non-zero. However, it is meaningless to discuss the convergence at zero, which is not always the result that we want for NMF. Thus, we theoretically illustrate the advantages of the cAG algorithm.

Keywords: Non-negative matrix factorizations, convergence, cAG algorithm, equilibrium point, stability.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1697

Authors: Meredith Stevenson, John E. Porter

Abstract:

Since the pioneering work of Zadeh, fuzzy set theory has been applied to a myriad of areas. Song and Chissom introduced the concept of fuzzy time series and applied some methods to the enrollments of the University of Alabama. In recent years, a number of techniques have been proposed for forecasting based on fuzzy set theory methods. These methods have either used enrollment numbers or differences of enrollments as the universe of discourse. We propose using the year to year percentage change as the universe of discourse. In this communication, the approach of Jilani, Burney, and Ardil is modified by using the year to year percentage change as the universe of discourse. We use enrollment figures for the University of Alabama to illustrate our proposed method. The proposed method results in better forecasting accuracy than existing models.

Keywords: Fuzzy forecasting, fuzzy time series, fuzzified enrollments, time-invariant model

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2504

Authors: Paul O'Leary, Matthew Harker

Abstract:

This paper presents unified theory for local (Savitzky- Golay) and global polynomial smoothing. The algebraic framework can represent any polynomial approximation and is seamless from low degree local, to high degree global approximations. The representation of the smoothing operator as a projection onto orthonormal basis functions enables the computation of: the covariance matrix for noise propagation through the filter; the noise gain and; the frequency response of the polynomial filters. A virtually perfect Gram polynomial basis is synthesized, whereby polynomials of degree d = 1000 can be synthesized without significant errors. The perfect basis ensures that the filters are strictly polynomial preserving. Given n points and a support length ls = 2m + 1 then the smoothing operator is strictly linear phase for the points xi, i = m+1. . . n-m. The method is demonstrated on geometric surfaces data lying on an invariant 2D lattice.

Keywords: Gram polynomials, Savitzky-Golay Smoothing, Discrete Polynomial Moments

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2790