A Fast Nonlinear Model Identification Method

The identification of nonlinear dynamic systems using linear-in-the-parameters models is studied. A fast recursive algorithm (FRA) is proposed to select both the model structure and to estimate the model parameters. Unlike orthogonal least squares (OLS) method, FRA solves the least-squares problem recursively over the model order without requiring matrix decomposition. The computational complexity of both algorithms is analyzed, along with their numerical stability. The new method is shown to require much less computational effort and is also numerically more stable than OLS.

Difference equation sensor characterization algorithms for two-thermocouple probes

The characterization of thermocouple sensors for temperature measurement in variable flow environments is a challenging problem. In this paper, novel difference equation-based algorithms are presented that allow in situ characterization of temperature measurement probes consisting of two-thermocouple sensors with differing time constants. Linear and non-linear least squares formulations of the characterization problem are introduced and compared in terms of their computational complexity, robustness to noise and statistical properties. With the aid of this analysis, least squares optimization procedures that yield unbiased estimates are identified. The main contribution of the paper is the development of a linear two-parameter generalized total least squares formulation of the sensor characterization problem. Monte-Carlo simulation results are used to support the analysis.

Approach to automatic video motion segmentation

A novel, fast automatic motion segmentation approach is presented. It differs from conventional pixel or edge based motion segmentation approaches in that the proposed method uses labelled regions (facets) to segment various video objects from the background. Facets are clustered into objects based on their motion and proximity details using Bayesian logic. Because the number of facets is usually much lower than the number of edges and points, using facets can greatly reduce the computational complexity of motion segmentation. The proposed method can tackle efficiently the complexity of video object motion tracking, and offers potential for real-time content-based video annotation.

A new Jacobian matrix for optimal learning of single-layer neural networks

This paper investigates the learning of a wide class of single-hidden-layer feedforward neural networks (SLFNs) with two sets of adjustable parameters, i.e., the nonlinear parameters in the hidden nodes and the linear output weights. The main objective is to both speed up the convergence of second-order learning algorithms such as Levenberg-Marquardt (LM), as well as to improve the network performance. This is achieved here by reducing the dimension of the solution space and by introducing a new Jacobian matrix. Unlike conventional supervised learning methods which optimize these two sets of parameters simultaneously, the linear output weights are first converted into dependent parameters, thereby removing the need for their explicit computation. Consequently, the neural network (NN) learning is performed over a solution space of reduced dimension. A new Jacobian matrix is then proposed for use with the popular second-order learning methods in order to achieve a more accurate approximation of the cost function. The efficacy of the proposed method is shown through an analysis of the computational complexity and by presenting simulation results from four different examples.

A novel continuous forward algorithm for RBF neural modelling

A continuous forward algorithm (CFA) is proposed for nonlinear modelling and identification using radial basis function (RBF) neural networks. The problem considered here is simultaneous network construction and parameter optimization, well-known to be a mixed integer hard one. The proposed algorithm performs these two tasks within an integrated analytic framework, and offers two important advantages. First, the model performance can be significantly improved through continuous parameter optimization. Secondly, the neural representation can be built without generating and storing all candidate regressors, leading to significantly reduced memory usage and computational complexity. Computational complexity analysis and simulation results confirm the effectiveness.

SPECK-Based Lossless Multispectral Image Coding

This letter proposes an efficient extension of the set partitioning embedded block (SPECK) algorithm to lossless multispectral image coding. Such a wavelet-based coder is widely referred to in the literature, especially for lossless image coding, and is considered to be one of the most efficient techniques exhibiting very low computational complexity when compared with other state-of-the-art coders. The modification proposed in this letter is simple and provides significant improvement over the conventional SPECK. The key idea is to join each group of two consecutive wavelet-transformed spectral bands during the SPECK coding since they show high similarities with respect to insignificant sets at the same locations. Simulation results, carried out on a number of test images, demonstrate that this grouping procedure considerably saves on the bit budget for encoding the multispectral images.

Equalisation of time variant multipath channels using amplitude banded LMS algorithms

For the purpose of equalisation of rapidly time variant multipath channels, we derive a novel adaptive algorithm, the amplitude banded LMS (ABLMS); which implements a nonlinear adaptation based on a coefficient matrix. Then we develop the: ABLMS algorithm as the adaptation procedure for a linear transversal equaliser (LTE) and a decision feedback equaliser (DFE) where a parallel adaptation scheme is deployed. Computer simulations demonstrate that with a small increase of computational complexity, the ABLMS based parallel equalisers provide a significant improvement related to the conventional LMS DFE and the LMS LTE in the case of a second order Markov communication channel model.

A fast multi-output RBF neural network construction method

This paper investigates the center selection of multi-output radial basis function (RBF) networks, and a multi-output fast recursive algorithm (MFRA) is proposed. This method can not only reveal the significance of each candidate center based on the reduction in the trace of the error covariance matrix, but also can estimate the network weights simultaneously using a back substitution approach. The main contribution is that the center selection procedure and the weight estimation are performed within a well-defined regression context, leading to a significantly reduced computational complexity. The efficiency of the algorithm is confirmed by a computational complexity analysis, and simulation results demonstrate its effectiveness. (C) 2010 Elsevier B.V. All rights reserved.

Two-stage mixed discrete-continuous identification of radial basis function (RBF) neural models for nonlinear systems

The identification of nonlinear dynamic systems using radial basis function (RBF) neural models is studied in this paper. Given a model selection criterion, the main objective is to effectively and efficiently build a parsimonious compact neural model that generalizes well over unseen data. This is achieved by simultaneous model structure selection and optimization of the parameters over the continuous parameter space. It is a mixed-integer hard problem, and a unified analytic framework is proposed to enable an effective and efficient two-stage mixed discrete-continuous; identification procedure. This novel framework combines the advantages of an iterative discrete two-stage subset selection technique for model structure determination and the calculus-based continuous optimization of the model parameters. Computational complexity analysis and simulation studies confirm the efficacy of the proposed algorithm.

Inferring the conservative causal core of gene regulatory networks

Background
Inferring gene regulatory networks from large-scale expression data is an important problem that received much attention in recent years. These networks have the potential to gain insights into causal molecular interactions of biological processes. Hence, from a methodological point of view, reliable estimation methods based on observational data are needed to approach this problem practically.

Results
In this paper, we introduce a novel gene regulatory network inference (GRNI) algorithm, called C3NET. We compare C3NET with four well known methods, ARACNE, CLR, MRNET and RN, conducting in-depth numerical ensemble simulations and demonstrate also for biological expression data from E. coli that C3NET performs consistently better than the best known GRNI methods in the literature. In addition, it has also a low computational complexity. Since C3NET is based on estimates of mutual information values in conjunction with a maximization step, our numerical investigations demonstrate that our inference algorithm exploits causal structural information in the data efficiently.

Conclusions
For systems biology to succeed in the long run, it is of crucial importance to establish methods that extract large-scale gene networks from high-throughput data that reflect the underlying causal interactions among genes or gene products. Our method can contribute to this endeavor by demonstrating that an inference algorithm with a neat design permits not only a more intuitive and possibly biological interpretation of its working mechanism but can also result in superior results.

Comparing large graphs efficiently by margins of feature vectors

Measuring the structural similarity of graphs is a challenging and outstanding problem. Most of the classical approaches of the so-called exact graph matching methods are based on graph or subgraph isomorphic relations of the underlying graphs. In contrast to these methods in this paper we introduce a novel approach to measure the structural similarity of directed and undirected graphs that is mainly based on margins of feature vectors representing graphs. We introduce novel graph similarity and dissimilarity measures, provide some properties and analyze their algorithmic complexity. We find that the computational complexity of our measures is polynomial in the graph size and, hence, significantly better than classical methods from, e.g. exact graph matching which are NP-complete. Numerically, we provide some examples of our measure and compare the results with the well-known graph edit distance. (c) 2006 Elsevier Inc. All rights reserved.

Software-defined sphere decoding for FPGA-based MIMO detection

Sphere Decoding (SD) is a highly effective detection technique for Multiple-Input Multiple-Output (MIMO) wireless communications receivers, offering quasi-optimal accuracy with relatively low computational complexity as compared to the ideal ML detector. Despite this, the computational demands of even low-complexity SD variants, such as Fixed Complexity SD (FSD), remains such that implementation on modern software-defined network equipment is a highly challenging process, and indeed real-time solutions for MIMO systems such as 4 4 16-QAM 802.11n are unreported. This paper overcomes this barrier. By exploiting large-scale networks of fine-grained softwareprogrammable processors on Field Programmable Gate Array (FPGA), a series of unique SD implementations are presented, culminating in the only single-chip, real-time quasi-optimal SD for 44 16-QAM 802.11n MIMO. Furthermore, it demonstrates that the high performance software-defined architectures which enable these implementations exhibit cost comparable to dedicated circuit architectures.

Tree-Based Adaptive Spatial Detection for Adaptive Modulated MIMO Systems

Adaptive Multiple-Input Multiple-Output (MIMO) systems achieve a much higher information rate than conventional fixed schemes due to their ability to adapt their configurations according to the wireless communications environment. However, current adaptive MIMO detection schemes exhibit either low performance (and hence low spectral efficiency) or huge computational
complexity. In particular, whilst deterministic Sphere Decoder (SD) detection schemes are well established for static MIMO systems, exhibiting deterministic parallel structure, low computational complexity and quasi-ML detection performance, there are no corresponding adaptive schemes. This paper solves
this problem, describing a hybrid tree based adaptive modulation detection scheme. Fixed Complexity Sphere Decoding (FSD) and Real-Values FSD (RFSD) are modified and combined into a hybrid scheme exploited at low and medium SNR to provide the highest possible information rate with quasi-ML Bit Error
Rate (BER) performance, while Reduced Complexity RFSD, BChase and Decision Feedback (DFE) schemes are exploited in the high SNR regions. This algorithm provides the facility to balance the detection complexity with BER performance with compatible information rate in dynamic, adaptive MIMO communications
environments.

Invariant Information Local Sub-map Filter (IILSF) for Efficient Simultaneous Localisation and Mapping of Large Environments

This paper presents an Invariant Information Local Sub-map Filter (IILSF) as a technique for consistent Simultaneous Localisation and Mapping (SLAM) in a large environment. It harnesses the benefits of sub-map technique to improve the consistency and efficiency of Extended Kalman Filter (EKF) based SLAM. The IILSF makes use of invariant information obtained from estimated locations of features in independent sub-maps, instead of incorporating every observation directly into the global map. Then the global map is updated at regular intervals. Applying this technique to the EKF based SLAM algorithm: (a) reduces the computational complexity of maintaining the global map estimates and (b) simplifies transformation complexities and data association ambiguities usually experienced in fusing sub-maps together. Simulation results show that the method was able to accurately fuse local map observations to generate an efficient and consistent global map, in addition to significantly reducing computational cost and data association ambiguities.

A multi-output two-stage locally regularized model construction method using the extreme learning machine

This paper investigates the construction of linear-in-the-parameters (LITP) models for multi-output regression problems. Most existing stepwise forward algorithms choose the regressor terms one by one, each time maximizing the model error reduction ratio. The drawback is that such procedures cannot guarantee a sparse model, especially under highly noisy learning conditions. The main objective of this paper is to improve the sparsity and generalization capability of a model for multi-output regression problems, while reducing the computational complexity. This is achieved by proposing a novel multi-output two-stage locally regularized model construction (MTLRMC) method using the extreme learning machine (ELM). In this new algorithm, the nonlinear parameters in each term, such as the width of the Gaussian function and the power of a polynomial term, are firstly determined by the ELM. An initial multi-output LITP model is then generated according to the termination criteria in the first stage. The significance of each selected regressor is checked and the insignificant ones are replaced at the second stage. The proposed method can produce an optimized compact model by using the regularized parameters. Further, to reduce the computational complexity, a proper regression context is used to allow fast implementation of the proposed method. Simulation results confirm the effectiveness of the proposed technique. © 2013 Elsevier B.V.