Chi-Yung Lee

Applying a Functional Neurofuzzy Network to Real-Time Lane Detection and Front-Vehicle Distance Measurement

Most traffic accidents resulted from distraction, inattention to surrounding cars, and driving fatigue. In order to protect drivers, a real-time lane-detection and front-vehicle distance measurement system that uses a mounted camera inside a vehicle has been designed for safe driving. For lane detection, the lane-boundary information is derived from the fan-scanning-detection method. The system calculates the departure degree according to the angular relationship of the boundaries and sends a suitable warning signal to drivers. For front-vehicle distance measurement, we use the front vehicles shadow underneath it to identify the position of the front vehicle. The real distance is estimated by the use of the functional neurofuzzy network. The experimental results show that the system works successfully in real-time environment.

A self-constructing fuzzy CMAC model and its applications

Abstract This work presents a self-constructing fuzzy cerebellar model articulation controller (SC-FCMAC) model for various applications. A self-constructing learning algorithm, which consists of the self-clustering method (SCM) and the back-propagation algorithm, is presented. The proposed SCM scheme is a rapid, one-pass algorithm which dynamically estimates the number of hypercube cells in input data space. The clustering method does not require prior knowledge, such as the number of clusters in a data set. The back-propagation algorithm is applied to tune the adjustable parameters. Simulation results are obtained to show the performance and applicability of the proposed model.

A self-adaptive quantum radial basis function network for classification applications

A self-adaptive quantum radial basis function network (QRBFN) is proposed for classification applications. The QRBFN model is a three-layer structure. The hidden layer of the QRBFN model contains quantum function neurons, which are multilevel activation functions. Each quantum function neuron is composed of the sum of sigmoid functions shifted by quantum intervals. A self-adaptive learning algorithm, which consists of the self-clustering algorithm (SCA) and the backpropagation algorithm, is proposed. The proposed the SCA method is a fast, one-pass algorithm for a dynamic estimation of the number of clusters in an input data space. The backpropagation algorithm is used to tune the adjustable parameters. Simulation results were conducted to show the performance and applicability of the proposed model.

Pattern recognition using neural-fuzzy networks based on improved particle swam optimization

This paper introduces a recurrent neural-fuzzy network (RNFN) based on improved particle swarm optimization (IPSO) for pattern recognition applications. The proposed IPSO method consists of the modified evolutionary direction operator (MEDO) and the traditional PSO. A novel MEDO combining the evolutionary direction operator (EDO) and the migration operation is also proposed. Hence, the proposed IPSO method can improve the ability of searching global solution. Experimental results have shown that the proposed IPSO method has a better performance than the traditional PSO in the human body classification and the skin color detection.

A novel hybrid learning algorithm for parametric fuzzy CMAC networks and its classification applications

This paper shows fundamentals and applications of the parametric fuzzy cerebellar model articulation controller (P-FCMAC) network. It resembles a neural structure that derived from the Albus CMAC and Takagi-Sugeno-Kang parametric fuzzy inference systems. In this paper, a novel hybrid learning which consists of self-clustering algorithm (SCA) and modified genetic algorithms (MGA) is proposed for solving the classification problems. The SCA scheme is a fast, one-pass algorithm for a dynamic estimation of the number of hypercube cells in an input data space. The clustering technique does not require prior knowledge such as the number of clusters present in a data set. The number of fuzzy hypercube cells and the adjustable parameters in P-FCMAC are designed by the MGA. The MGA uses the sequential-search based efficient generation (SSEG) method to generate an initial population to determine the most efficient mutation points. Illustrative examples were conducted to show the performance and applicability of the proposed model.

A self-organizing recurrent fuzzy CMAC model for dynamic system identification

This paper presents a self-organizing recurrent fuzzy cerebellar model articulation controller (RFCMAC) model for identifying a dynamic system. The recurrent network is embedded in the RFCMAC by adding feedback connections with a receptive field cell to the RFCMAC, where the feedback units act as memory elements. A nonconstant differentiable Gaussian basis function is used to model the hypercube structure and the fuzzy weight. An online learning algorithm is proposed for the automatic construction of the proposed model during the learning procedure. The self-constructing input space partition is based on the degree measure to appropriately determine the various distributions of the input training data. A gradient descent learning algorithm is used to adjust the free parameters. The advantages of the proposed RFCMAC model are summarized as follows: (1) it requires much lower memory requirement than other models; (2) it selects the memory structure parameters automatically; and (3) it has better identification performance than other recurrent networks.

Using Least Squares Support Vector Machines for Adaptive Communication Channel Equalization

Adaptive equalizers are used in digital communication system receivers to mitigate the effects of non-ideal channel characteristics and to obtain reliable data transmission. In this paper, we adopt least squares support vector machines (LS-SVM) for adaptive communication channel equalization. The LS-SVM involves equality instead of inequality constraints and works with a least squares cost function. Since the complexity and computational time of a LS-SVM equalizer are less than an optimal equalizer, the LS-SVM equalizer is suitable for adaptive digital communication and signal processing applications. Computer simulation results show that the bit error rate of the LS-SVM equalizer is very close to that of the optimal equalizer and better than multilayer perceptron (MLP) and wavelet neural network (WNN) equalizers.

A functional neural fuzzy network for classification applications

Research highlights? A functional neural fuzzy system model is proposed. We adopts a functional neural network to the consequent part of the fuzzy rules. ? Orthogonal polynomials and linearly independent functions are used for a functional expansion of the functional neural network. ? An online learning algorithm, which consist of structure learning algorithm and parameter learning algorithm, is proposed. ? The average testing accuracy rates of the functional neural fuzzy system in Iris data and wine classification data were 98.1% and 99.1%. This study presents a functional neural fuzzy network (FNFN) for classification applications. The proposed FNFN model adopts a functional neural network (FLNN) to the consequent part of the fuzzy rules. Orthogonal polynomials and linearly independent functions are used for a functional expansion of the FLNN. Thus, the consequent part of the proposed FNFN model is a nonlinear combination of input variables. The FNFN model can construct its structure and adapt its free parameters with online learning algorithms, which consist of structure learning algorithm and parameter learning algorithm. The structure learning algorithm is based on the entropy measure to determine the number of fuzzy rules. The parameter learning algorithm, based on the gradient descent method, can adjust the shapes of the membership functions and the corresponding weights of the FLNN. Finally, the FNFN model is applied to various simulations. The simulation results for the Iris, Wisconsin breast cancer, and wine classifications show that FNFN model has superior performance than other models for classification applications.

An adaptive fuzzy predictor based handoff algorithm for heterogeneous network

In the future, the mobile data networks must consist of several tiers that would overlap with each other. The goal of this paper is to propose a new handoff algorithm for heterogeneous network. The heterogeneous network consists of cellular phone network (i.e., GPRS and UMTS) and IEEE 802.11 wireless local area network (WLAN). The proposed algorithm is an adaptive fuzzy predictor based handoff algorithm that can adapt with the dynamic conditions in the heterogeneous network. First, we use fuzzy predictor to predict the received signal strength (RSS) in cellular phone network and WLAN. Second, we propose a fuzzy decision algorithm to determine the possibility of handoff according to fuzzy decision algorithm, Our approach shows an efficient solution on the issue of the handoff problem in heterogeneous networks, Therefore, the effort needs to be devoted to finding new handoff strategies, and the impact on the novel fuzzy reasoning handoff strategy is necessary.

2D/3D Face Recognition Using Neural Networks Based on Hybrid Taguchi-Particle Swarm Optimization

We present a method of face recognition using facial texture and surface information. We first use Gabor wavelets extracting local features at different scales and orientations by gray facial images, then combine the texture with the surface feature vectors based on principal component analysis (PCA) to obtain feature vectors from gray and facial surface images. We propose a hybrid Taguchi particle swarm optimization (HTPSO) algorithm for face recognition based on multilayer neural networks as an identification model. Experimental results demonstrate the efficiency of our method for different face poses and facial expressions. In addition, our work compared with other proposed approaches such as back-propagation (BP), particle swarm optimization (PSO) and the genetic algorithm (GA). With different data modality the experimental results demonstrated that the proposed HTPSO learning algorithm is better than the other approaches in recognition rates. The texture and shape features can improve the efficiency of face recognition.