System Maintenance:
There may be intermittent impact on performance while updates are in progress. We apologize for the inconvenience.
By Topic

Engineering of Intelligent Systems, 2006 IEEE International Conference on

Date 0-0 2006

Filter Results

Displaying Results 1 - 25 of 99
  • Message from the General Chairman

    Publication Year: 2006 , Page(s): 1
    Save to Project icon | Request Permissions | PDF file iconPDF (106 KB) |  | HTML iconHTML  
    Freely Available from IEEE
  • Contributor Listings

    Publication Year: 2006 , Page(s): 1 - 2
    Save to Project icon | Request Permissions | PDF file iconPDF (73 KB)  
    Freely Available from IEEE
  • Table of contents

    Publication Year: 2006 , Page(s): 1 - 9
    Save to Project icon | Request Permissions | PDF file iconPDF (4833 KB)  
    Freely Available from IEEE
  • Non-Linear Predictors based on the Functionally Expanded Neural Networks for Speech Feature Extraction

    Publication Year: 2006 , Page(s): 1 - 5
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (4847 KB) |  | HTML iconHTML  

    In this paper we focus on the design of the feature extractor stage of the speech recognition system which aims to compute optimal vectors for the next phoneme classification stage. We propose a new non-linear feature extraction method based on the linear-in-parameters functionally expanded neural network (FENN) model. The main idea is to design an improved and flexible feature extractor which can effectively account for some of the significant non-linear phenomena usually observed in the speech production process. The effectiveness of the proposed method is assessed on phoneme classification tasks. Specifically, we evaluate the performances on the telephone quality NTIMIT database, focusing the investigations on highly confusable phonemes such as front vowels: /ih/, /ey/, /eh/, /ae/. The results are compared with other widely used coding methods namely, the linear predictive coding (LPC) and the Mel frequency cepstral coding (MFCC). The experiments show a relative improvement in the rates through the use of our proposed non-linear feature extractor technique View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Estimation and Decision Fusion: A Survey

    Publication Year: 2006 , Page(s): 1 - 6
    Cited by:  Papers (1)  |  Patents (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (619 KB) |  | HTML iconHTML  

    Data fusion has been applied to a large number of fields and the corresponding applications utilize numerous mathematical tools. This survey limits the scope to some aspects of estimation and decision fusion. In estimation fusion our main focus is on the cross-correlation between local estimates from different sources. On the other hand, the problem of decision fusion is discussed with emphasis on the classifier combining techniques View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Complex Stochastic Systems Modeling and Control via Iterative Machine Learning

    Publication Year: 2006 , Page(s): 1 - 6
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (3885 KB) |  | HTML iconHTML  

    Complex stochastic systems require the control of their stochastic distributions. This keynote paper addresses both modelling and control of such systems and consists of the following aspects: 1) neural network based modelling of the stochastic distribution systems; 2) control framework for the stochastic profile control of the systems; 3) iterative learning of the space variables so as to achieve a batch-by-batch improvement of the closed loop performance. The above includes our originated research on complex stochastic systems in terms of probability density function (pdfs) control, where neural networks such as RBF was used to approximate the output pdfs and the system dynamics. This was followed by the iterative machine learning for the RBF basis functions on a batch-by-batch basis so as to improve the closed loop performance both in the time and in the space. Applications to particle size distribution control and 3D paper Web distribution control was discussed in the presentation View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Linking Human and Machine Brains

    Publication Year: 2006 , Page(s): 1 - 6
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2067 KB) |  | HTML iconHTML  

    A look is taken here at how the use of implant technology is rapidly diminishing the effects of certain neural illnesses and distinctly increasing the range of abilities of those affected. An indication is given of a number of problem areas in which such technology has already had a profound effect, a key element being the need for a clear interface linking the human brain directly with a computer. In order to assess the possible opportunities, both human and animal studies are reported on. The main thrust of the paper is however a discussion of neural implant experimentation linking the human nervous system bi-directionally with the Internet. With this in place neural signals were transmitted to various technological devices to directly control them, in some cases via the Internet, and feedback to the brain was obtained from such as the fingertips of a robot hand, ultrasonic (extra) sensory input and neural signals directly from another human's nervous system. Consideration is given to the prospects for neural implant technology in the future, both in the short term as a therapeutic device and in the long term as a form of enhancement, including the realistic potential for thought communication -potentially opening up commercial opportunities. Clearly though, an individual whose brain is part human - part machine can have abilities that far surpass those with a human brain alone. Will such an individual exhibit different moral and ethical values to those of a human? If so, what effects might this have on society? View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fuzzy-Rule based Load Pattern Classifier for Short-Tern Electrical Load Forecasting

    Publication Year: 2006 , Page(s): 1 - 6
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (4759 KB) |  | HTML iconHTML  

    Based on the knowledge of historical data sets, a fuzzy rule-based classifier for electrical load pattern classification is set up. Considering with the accuracy and interpretation of fuzzy rules, multi-objective genetic algorithm are applied to choose the Pareto optimum rules that are used to classify electrical load. In the computation experiments, the generated fuzzy rule-based classifier is used to load forecasting, the computation results show that it leads to high classification performance, and it can supply more sufficient and effective historical data for load forecasting, better performance of load forecasting is gained accordingly View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Neural Nets for On-line Isolated Handwritten Character Recognition: A Comparative Study

    Publication Year: 2006 , Page(s): 1 - 6
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (3930 KB) |  | HTML iconHTML  

    Handwriting processing is a domain in great expansion which begins to see several industrial realizations. The field of personal computing has begun to make a transition from the desktop to handheld devices, thereby requiring input paradigms that are more suited for single hand entry than a keyboard. Online handwriting recognition allows for such input modalities. Handwriting recognition has always been a tough problem because of the handwriting variability, ambiguity and illegibility. This paper describes a simple approach involved in online handwriting recognition. Conventionally, the data obtained needs a lot of preprocessing including filtering, smoothing, slant removing and size normalization before recognition process. Instead of doing such lengthy preprocessing, this paper presents a simple approach to extract the useful character information. The whole process requires no preprocessing and size normalization. The method is applicable for off-line character recognition as well. This is a writer-independent system based on two neural net (NN) techniques: back propagation neural network (BPN) and counter propagation neural network (CPN). Performances of BPN and CPN are tested for upper-case English alphabets for a number of different styles from different peoples View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • In-building Localization using Neural Networks

    Publication Year: 2006 , Page(s): 1 - 6
    Cited by:  Papers (8)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (4153 KB) |  | HTML iconHTML  

    Location awareness is key capability of context-aware ubiquitous environments. Received signal strength (RSS) based localization is increasingly popular choice especially for indoor scenarios after pervasive adoption of IEEE 802.11 wireless LAN. Fundamental requirement of such localization systems is to estimate location from RSS at a particular location. Multi-path propagation effects make RSS to fluctuate in unpredictable manner, introducing uncertainty in location estimation. Moreover, in real life situations RSS values are not available at some locations all the time making the problem more difficult. We employ modular multi-layer perceptron (MMLP) approach to effectively reduce the uncertainty in location estimation system. It provides better location estimation results than other approaches and systematically caters for unavailable signals at estimation time View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Rotation-Invariant Features for Texture Image Classification

    Publication Year: 2006 , Page(s): 1 - 4
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1056 KB) |  | HTML iconHTML  

    Texture features based on wavelet transform are sensitive to texture rotation and translation. This paper develops a new rotation invariant texture analysis technique using principal components analysis (PCA) and wavelet transform. The PCA is first used to calculate the angle of the principal direction of the texture. Then, the texture is rotated in the opposite direction by the same angle as detected by PCA. Finally a wavelet transform is applied to the preprocessed texture to extract features which are rotation invariant View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Blind Equalization of Communication Channels with Equal Energy Sources Using a Combined HOS-SOS Approach

    Publication Year: 2006 , Page(s): 1 - 3
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (262 KB) |  | HTML iconHTML  

    This paper presents a combined higher order statistics (HOS) and second order statistics (SOS) approach to blind equalization of white as well as colored sources. Remarkable convergence speed has been achieved through an additional term in the cost function based on the energies of the source symbols. Computational complexity of the new algorithm is less than the previously proposed hybrid approach and exhibits enhanced robustness compared against low SNR. The algorithm is valid for equal energy sources and provides convergence in only one epoch (training cycle) View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Computation Process Evolution

    Publication Year: 2006 , Page(s): 1 - 6
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2707 KB) |  | HTML iconHTML  

    Unlike other genetic methods which are devoted to optimize the input data, this paper proposes an approach, CPE, aiming at finding the computation process of any problem by only using a few input and output data, consisting of the cases needed to be satisfied and those needed to be avoided. It first encodes the antibody using the method similar to that of gene expression programming (GEP), a new efficient technique of genetic programming (GP) with linear representation. Through the gradual evolution, the affinity between antibody and the non-selves become more and more intense. At the same time, every time after the chromosomes are mutated, the chromosomes should be checked to determine whether the antibody chromosome would match the selves, which are the conditions that should be satisfied. Two kind of experiment are examined in order to test the performance of the approach. The results show that CPE evolves out the data-processing processes which are exactly the same as those from which the experimental input data were generated, and compared with GP and GEP which is currently one of the most efficient genetic methods, CPE experiences shorter evolution process. Most importantly, unlike previous evolutionary methods that only consider increasing fitness, this approach takes into account both the goal (fitness) and the constraints of actual problems, which makes it possible to solve complex real problems using evolutionary computation View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Neuro-Predictive Control for Automotive Air Conditioning System

    Publication Year: 2006 , Page(s): 1 - 6
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (4089 KB) |  | HTML iconHTML  

    This paper presents a neuro-predictive controller for temperature control of automotive air conditioning system. A numerical model for the automotive refrigeration cycle, which includes transient operating conditions, has been employed in simulations. In this model, which has been created from numerous laboratory tests on a typical passenger car, the vehicle is divided into two linked modules representing the air conditioning (A/C) system as well as the passenger compartment climate. Moreover, the thermal loads have been considered. This system demonstrates variable and large time delays, which is the case in reality. The simulation results show good performance of the proposed controller View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fuzzy Models for Time Series Analysis: Towards Systematic Data Pre-processing

    Publication Year: 2006 , Page(s): 1 - 5
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (3649 KB) |  | HTML iconHTML  

    Time series forecasting is an all-pervasive task that affects almost all disciplines. Given that time varied phenomena almost invariably need pre-processing, it is important to develop a framework where such pre-processing is executed in a systematic and transparent manner. In this paper, we investigate the effect of data pre-processing on the forecast performance of subtractive clustering fuzzy model. Our work on benchmark data sets (US Census Board and US Federal Reserve data) shows that ad hoc application of pre-processing techniques is not optimal. We have used autocorrelation functions to understand both the behavior of time series and the effects of different preprocessing methods on prediction accuracy. Our results indicate that the use of autocorrelation functions to determine the suitability of different pre-processing methods is beneficial View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Reasoning Based on Rules Extracted from Trained Neural Networks via Formal Concept Analysis

    Publication Year: 2006 , Page(s): 1 - 6
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (5758 KB) |  | HTML iconHTML  

    Due to their capability of dealing with nonlinear problems, artificial neural networks (ANN) are widely used with several purposes. Once trained, they are also capable of solving unprecedented situations, keeping tolerable errors in their outputs. However, ANN are considered essentially "black boxes". Therefore, humans can not assimilate the knowledge kept by those nets, since such knowledge is implicitly represented by their connection weights. In this paper, a new approach to extract knowledge rules from ANN previously trained through formal concept analysis is presented. The method allows to the knowledge engineer understand the industrial process that is being analyzed, through implications rules of the type if... then. As an example of application a solar energy system is considered. The rules obtained are validated through an expert domain View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Output Feedback Decentralized Control of Multi-Agent Manipulation Systems

    Publication Year: 2006 , Page(s): 1 - 7
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2011 KB) |  | HTML iconHTML  

    In this paper, decentralized control algorithms for cooperative multi-agent manipulation systems are developed. To control the positions of the agents and the exerted forces on the object in the presence of uncertainties in the dynamics of the agents, two different methods are considered. In the first approach, robust control of the system is proposed. Using the Lyapunov stability method, the convergence of the position errors to zero is demonstrated. Also a bound on the errors of the exerted forces is achieved which can be made small enough by choosing the appropriate estimated values for the physical parameters of the agents. In the second approach, adaptive control algorithm is proposed and the convergence of both the tracking errors and the errors of the exerted forces to zero is guaranteed. To avoid the difficulties of using the velocity sensors within the cooperative system, an output feedback scheme using linear observers is also proposed View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Enhancing the Security of Intelligent Transportation Systems (ITS) using Iris/Finger-based Multimodal Biometrics

    Publication Year: 2006 , Page(s): 1 - 6
    Cited by:  Papers (1)  |  Patents (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (3887 KB) |  | HTML iconHTML  

    Intelligent transportation system (ITS) is a term used for a wide range of technologies incorporated into traditional transportation infrastructure and vehicles, these include communication and information technologies such as traffic cameras, ramp meters, variable message signs (VMS), traffic control centers (TCC), accident avoidance technologies, incident management and more, that promises to maximize transportation safety and efficiency. Due to the complex architecture of ITS and its subsystems, the security and safety of ITS to operate efficiently is becoming an emerging issue. Slight error or deliberate changes to the one of the components of ITS may disrupt the whole system and may create dangerous roadway conditions. For these reasons, the conventional password protection techniques used in ITS control systems must be converted into more secure biometric systems. In this paper we present multimodal iris/fingerprint biometric system and its application in ITS. The performance evaluation model is presented to check the efficiency of the system View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Semiparametric Model Approach to Financial Bankruptcy Prediction

    Publication Year: 2006 , Page(s): 1 - 6
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (5939 KB) |  | HTML iconHTML  

    In this study, we propose a model that achieves both accurate modeling and sustainable model stability for corporate bankruptcy prediction. This model is to model the given samples accurately as well as to respond adequately to the unknown inputs by employing semiparametric approach where parametric model and nonparametric neural networks (NNs) are combined. By exploring the structural relationships within the available sample data, the proposed model is assumed to retain the advantages of both parametric and nonparametric models. The proposed model is compared to pure parametric models such as multivariate discriminant analysis (MDA) and logistic regression (LR), and pure nonparametric model such as NNs. Each model predicts the default probability of a company and classifies the company into an appropriate group as either bankrupt or healthy. Experimental results demonstrated that the proposed semiparametric model showed superior performance in terms of model stability and prediction accuracy in bankruptcy prediction View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Texture Classification Using Wavelet Frame Representation Based Feature

    Publication Year: 2006 , Page(s): 1 - 4
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1993 KB) |  | HTML iconHTML  

    Texture classification is an important component in image analysis and understanding. The wavelet, a multiresolution signal analysis technique, has been successfully applied to describe the texture. It is noticed that the wavelet transform modulus maximum and minimum can effectively characterize a signal. This paper employs the density of modulus maxima and the density of modulus minima of the wavelet frame representation as features for texture classification. In order to avoid the problem of curse of dimensionality, the feature selection algorithm, sequential forward floating selection (SFFS), is used to select feature subset. The experimental results on two benchmark databases indicate that the new feature is better than existing features based on modulus extrema, zero-crossings and local extrema View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Video based Parallel Face recognition using Gabor filter on homogeneous distributed systems

    Publication Year: 2006 , Page(s): 1 - 5
    Cited by:  Papers (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2301 KB) |  | HTML iconHTML  

    This research aimed at building a fast video, parallel face recognition system based on the well known Gabor filtering approach. Face recognition is done after face detection in each frame of the video, individually. The master-slave technique is employed as the parallel computing model. Each frame is processed by different slave personal computers (PC) attached to the master, which acquire and distribute frames. It is believed that this approach can be used for practical face recognition applications with some further optimization View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • New Hybrid Control Approach for Cascade Temperature Control in Catalyst Regeneration Process

    Publication Year: 2006 , Page(s): 1 - 6
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (3454 KB) |  | HTML iconHTML  

    This paper describes an improvement in performance achieved over an existing feedback control system for the temperature control of catalyst bed in chemical reactors. The improvements are focused on devising a new control technique that improves the system behavior against any kind of disturbances. Catalyst regeneration process regenerates the catalyst of a chemical reactor. An important variable, which is to be controlled, is the temperature of catalyst bed. A very high temperature may destroy the qualities of the catalyst, while a low temperature may result in long burning time. The existing temperature control system in use is designed around conventional simple feedback control loop with proportional type of controller. The feedback control technique imposes various limitations like poor time response and relative stability. The control strategy suggested in this paper, involves the use of multiple cascade control loop with advanced fuzzy knowledge based controller and the feed-forward control. The use of two feedback loops in cascade, divides the total time lag present in the system into two parts. This paper details the verification of the comparison between the performance of an existing simple feedback control and an advanced cascade control technique based on fuzzy logic controller, for the catalyst regeneration process at modern continuous process steel plant. The improvement in system performance based on cascade control approach over the conventional one from the point of view of improved time response and relative stability is verified through simulation results View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Formal Concept Analysis for Data Mining: Theoretical and Practical Approaches

    Publication Year: 2006 , Page(s): 1 - 6
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (5377 KB) |  | HTML iconHTML  

    Knowledge discovery in databases (KDD) is the most widely known process with the purpose of knowledge extraction. Formal concept analysis (FCA) is proposed here as an alternative step in KDD process, due to its capacity of generating diagrams that facilitate data representation and analysis. FCA can perform the task of data mining (DM), supporting users in knowledge management. Both theoretical and practical aspects are presented here View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Innovative Inverse Control Techniques for Adaptive Tracking of Nonlinear Dynamic Plants

    Publication Year: 2006 , Page(s): 1 - 6
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1529 KB) |  | HTML iconHTML  

    Adaptive tracking of nonlinear dynamic plants is an active area of research. The main difficulty felt in establishing the tracking of nonlinear dynamic plants is the computational complexity in controller design. This paper presents novel adaptive tracking techniques for a class of nonlinear dynamic plants based on a control oriented model known as U-model. The use of U-model reduces the computational complexity of the controller design that occurs when using other modelling frame works such as NARMAX (nonlinear autoregressive moving average with exogenous inputs) model. Four different adaptive control techniques based on U-model are discussed namely adaptive inverse control (AIC), adaptive internal model control (AIMC), AIC with pole-placement and AIMC with pole-placement. The proposed techniques are implemented in real-time on a laboratory scale experimental setup for speed control of DC-motor. The results of the experiments are provided View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Co-evolutionary Competitive Multi-expert Approach to Image Compression with Neural Networks

    Publication Year: 2006 , Page(s): 1 - 5
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (3181 KB) |  | HTML iconHTML  

    Bottle-neck MLP neural networks have been used in image compression and a few methods are developed to increase the compression quality. In this paper, a new co-evolutionary method is proposed to further improve the compression efficiency. A heterogeneous set of networks co-evolve and compete to compress different parts of an image with different characteristics. The results indicate a great improvement over the non-evolving methods View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.