Speaker identification using multi-step clustering algorithm with transformation-based GMM

To improve the performance of speaker recognition, the embedded linear transformation is used to integrate both transformation and diagonal-covariance Caussian mixture into a unified framework. In the case, the mixture number of GMM must be fixed in model training. The cluster expectation-maximization (EM) algorithm is a well-known technique in which the mixture number is regarded as an estimated parameter. This paper presents a new model structure that integrates a multi-step cluster algorithm into the estimating process of GMM with the embedded transformation. In the approach, the transformation matrix, the mixture number and model parameters are simultaneously estimated according to a maximum likelihood criterion. The proposed method is demonstrated on a database of three data sessions for text independent speaker identification. The experiments show that this method outperforms the traditional GMM with cluster EM algorithm. This text was submitted by the authors in English.     

Data fitting with a spline using a real-coded genetic algorithm

To obtain a good approximation for data fitting with a spline, frequently we have to deal with knots as variables. The problem to be solved then becomes a continuous nonlinear and multivariate optimization problem with many local optima. Therefore, it is difficult to obtain the global optimum. In this paper, we propose a method for solving this problem by using a real-coded genetic algorithm. Our method can treat not only data with a smooth underlying function, but also data with an underlying function having discontinuous points and/or cusps. We search for the best model among candidate models by using the Bayes Information Criterion (BIC). With this, we can appropriately determine the number and locations of knots automatically and simultaneously. Five examples of data fitting are given to show the performance of our method.     

A fuzzy GARCH model applied to stock market scenario using a genetic algorithm

In this paper, we derive a new application of fuzzy systems designed for a generalized autoregression conditional heteroscedasticity (GARCH) model. In general, stock market performance is time-varying and nonlinear, and exhibits properties of clustering. The latter means simply that certain large changes tend to follow other large changes, and in general small changes tend to follow other small changes. This paper shows results from using the method of functional fuzzy systems to analyze the clustering in the case of a GARCH model.The optimal parameters of the fuzzy membership functions and GARCH model are extracted using a genetic algorithm (GA). The GA method aims to achieve a global optimal solution with a fast convergence rate for this fuzzy GARCH model estimation problem. From the simulation results, we have determined that the performance is significantly improved if the leverage effect of clustering is considered in the GARCH model. The simulations use stock market data from the Taiwan weighted index (Taiwan) and the NASDAQ composite index (NASDAQ) to illustrate the performance of the proposed method.     

The model-based dynamic hand posture identification using genetic algorithm

This paper proposes a new hand posture identification system which applies genetic algorithm to develop an efficient 3D hand-model-fitting method. The 3D hand-model-fitting method consists of (1) finding the closed-form inverse kinematics solution, (2) defining the alignment measure function for the wrist-fitting process, and (3) applying genetic algorithm to develop the dynamic hand posture identification process. In contrast to the conventional computationally intensive hand-model-fitting methods, we develop an off-line training process to find the closed-form inverse kinematics solution functions, and a fast model-based hand posture identification process. In the experiments, we will illustrate that our hand posture identification system is very effective. Received: 10 April 1997 / Accepted: 18 June 1998     

A genetic clustering algorithm using a message-based similarity measure

In this paper, a genetic clustering algorithm is described that uses a new similarity measure based message passing between data points and the candidate centers described by the chromosome. In the new algorithm, a variable-length real-value chromosome representation and a set of problem-specific evolutionary operators are used. Therefore, the proposed GA with message-based similarity (GAMS) clustering algorithm is able to automatically evolve and find the optimal number of clusters as well as proper clusters of the data set. Effectiveness of GAMS clustering algorithm is demonstrated for both artificial and real-life data set. Experiment results demonstrated that the GAMS clustering algorithm has high performance, effectiveness and flexibility.     

Driver fatigue detection using a genetic algorithm

Nowadays, many traffic accidents occur due to driver fatigue. Driver fatigue detection based on computer vision is one of the most hopeful applications of image recognition technology. There are several factors that reflect driver's fatigue. Many efforts have been made to develop fatigue monitoring, but most of them focus on only a single behavior, a feature of the eyes, or a head motion, or mouth motion, etc. When fatigue monitoring is implemented on a real model, it is difficult to predict the driver fatigue accurately or reliably based only on a single driver behavior. Additionally, the changes in a driver's performance are more complicated and not reliable. In this article, we represent a model that simulates a space in a real car. A web camera as a vision sensor is located to acquire video-images of the driver. Three typical characteristics of driver fatigue are involved, pupil shape, eye blinking frequency, and yawn frequency. As the influences of these characteristics on driver fatigue are quite different from each other, we propose a genetic algorithm (GA)-based neural network (NN) system to fuse these three parameters. We use the GA to determine the structure of the neural network system. Finally, simulation results show that the proposed fatigue monitoring system detects driver fatigue probability more exactly and robustly. This work was presented in part at the 11th International Symposium on Artificial Life and Robotics, Oita, Japan, January 23–25, 2006     

Surface registration using a dynamic genetic algorithm

Robust and fast free-form surface registration is a useful technique in various areas such as object recognition and 3D model reconstruction for animation. Notably, an object model can be constructed, in principle, by surface registration and integration of range images of the target object from different views. In this paper, we propose to formulate the surface registration problem as a high dimensional optimization problem, which can be solved by a genetic algorithm (GA) (Genetic Algorithms in Search Optimization and Machine Learning, Addison-Wesley, Reading, MA, 1989). The performance of the GA for surface registration is highly dependent on its speed in evaluating the fitness function. A novel GA with a new fitness function and a new genetic operator is proposed. It can compute an optimal registration 1000 times faster than a conventional GA. The accuracy, speed and the robustness of the proposed method are verified by a number of real experiments.     

Distributed genetic algorithm for Gaussian mixture model based speaker identification

This paper presents a novel algorithm for reducing the computational complexity of identifying a speaker within a Gaussian mixture speaker model (GMM) framework. We have combined distributed genetic algorithm (DGA) and the Markov random field (MRF) to avoid typical local minima for speaker vector quantization. To improve the computation efficiency, only unstable chromosomes corresponding to speaker data parts are evolved. Identification accuracies of 93% were achieved for 100 Mandarin speakers.     

DNA fragment assembly using a grid-based genetic algorithm

In this paper we propose a genetic algorithm (GA) for solving the DNA fragment assembly problem in a computational grid. The algorithm, which is named GrEA, is a steady-state GA which uses a panmitic population, and it is based on computing parallel function evaluations in an asynchronous way. We have implemented GrEA on top of the Condor system, and we have used it to solve the DNA assembly problem. This is an NP-hard combinatorial optimization problem which is growing in importance and complexity as more research centers become involved on sequencing new genomes. While previous works on this problem have usually faced 30 K base pairs (bps) long instances, we have tackled here a 77 K bps long one to show how a grid system can move research forward. After analyzing the basic grid algorithm, we have studied the use of an improvement method to still enhance its scalability. Then, by using a grid composed of up to 150 computers, we have achieved time reductions from tens of days down to a few hours, and we have obtained near optimal solutions when solving the 77 K bps long instance (773 fragments). We conclude that our proposal is a promising approach to take advantage of a grid system to solve large DNA fragment assembly problem instances and also to learn more about grid metaheuristics as a new class of algorithms for really challenging problems.     

Analysis and validation of a simplified model of a wet flue gas cleaning system

Combustion of waste produces large variations in flue gas composition. These variations lead to dynamic effects in the emission control system. Variations in scrubber liquid pH is a typical effect, influencing the absorption of pollutants. A dynamic model for a wet system for SO₂ removal was built using an existing plant in Hobro, Denamrk as a model system. An experimental validation showed that the model has a good agreement with operating data. There was one important limitation to model performance. The absorption of HCl were not modelled and the experiments proved this to be an important process characteristic. A linear analysis of the model showed that the process is approximately a first order system. The process gain is however changing with changing chemical composition of the scrubber liquid, which makes a robust controller design hard to achieve. Some concluding remarks are made on plant and control system design.     

A mathematical model and a genetic algorithm for two-sided assembly line balancing

A two-sided assembly line is a type of production line where tasks are performed in parallel at both sides of the line. The line is often found in producing large products such as trucks and buses. This paper presents a mathematical model and a genetic algorithm (GA) for two-sided assembly line balancing (two-ALB). The mathematical model can be used as a foundation for further practical development in the design of two-sided assembly lines. In the GA, we adopt the strategy of localized evolution and steady-state reproduction to promote population diversity and search efficiency. When designing the GA components, including encoding and decoding schemes, procedures of forming the initial population, and genetic operators, we take account of the features specific to two-ALB. Through computational experiments, the performance of the proposed GA is compared with that of a heuristic and an existing GA with various problem instances. The experimental results show that the proposed GA outperforms the heuristic and the compared GA.     

Unit commitment problem with ramp rate constraint using a binary-real-coded genetic algorithm

The unit commitment problem (UCP) is a nonlinear mixed-integer optimization problem, encountered as one of the toughest problems in power systems. The problem becomes even more complicated when dynamic power limit based ramp rate constraint is taken into account. Due to the inadequacy of deterministic methods in handling large-size instances of the UCP, various metaheuristics are being considered as alternative algorithms to realistic power systems, among which genetic algorithm (GA) has been investigated widely since long back. Such proposals have been made for solving only the integer part of the UCP, along with some other approaches for the real part of the problem. Moreover, the ramp rate constraint is usually discussed only in the formulation part, without addressing how it could be implemented in an algorithm. In this paper, the GA is revisited with an attempt to solve both the integer and real parts of the UCP using a single algorithm, as well as to incorporate the ramp rate constraint in the proposed algorithm also. In the computational experiment carried out with power systems up to 100 units over 24-h time horizon, available in the literature, the performance of the proposed GA is found quite satisfactory in comparison with the previously reported results.     

A closed-loop logistic model with a spanning-tree based genetic algorithm

Due to the problem of global warming, the green supply chain management, in particular, closed-loop logistics, has drawn the attention of researchers. Although there were logistics models that were examined in the literatures, most of them were case based and not in a closed-loop. Therefore, they lacked generality and could not serve the purposes of recycling, reuse and recovery required in a green supply chain. In this study, the integration of forward and reverse logistics was investigated, and a generalized closed-loop model for the logistics planning was proposed by formulating a cyclic logistics network problem into an integer linear programming model. Moreover, the decisions for selecting the places of manufactories, distribution centers, and dismantlers with the respective operation units were supported with the minimum cost. A revised spanning-tree based genetic algorithm was also developed by using determinant encoding representation for solving this NP model. Numerical experiments were presented, and the results showed that the proposed model and algorithms were able to support the logistic decisions in a closed-loop supply chain efficiently and accurately.

Statement of scope and purposes

This study concerns with operations of 3R in the green supply chain logistics and the location selection optimization. Based on ‘cradle to cradle’ principle of a green product, a “closed-loop” structure of a network was proposed in order to integrate the environmental issues into a traditional logistic system. Due to NP-hard nature of the model, a Genetic Algorithm, which is based on spanning tree structure was developed. Test problems from the small size for accuracy to the large scale for efficiency have been demonstrated with comparison. The promising results have shown the applicability of the proposed model with the solution procedure.     

Structural topology optimization using a genetic algorithm with a morphological geometric representation scheme

This paper describes a versatile methodology for solving topology design optimization problems using a genetic algorithm (GA). The key to its effectiveness is a geometric representation scheme that works by specifying a skeleton which defines the underlying topology/connectivity of a structural continuum together with segments of material surrounding the skeleton. The required design variables are encoded in a chromosome which is in the form of a directed graph that embodies this underlying topology so that appropriate crossover and mutation operators can be devised to recombine and help preserve any desirable geometry characteristics of the design through succeeding generations in the evolutionary process. The overall methodology is first tested by solving ‘target matching’ problems—simulated topology optimization problems in each of which a ‘target’ geometry is first created and predefined as the optimum solution, and the objective of the optimization problem is to evolve design solutions to converge towards this target shape. The methodology is then applied to design two path-generating compliant mechanisms—large-displacement flexural structures that undergo some desired displacement paths at some point when given a straight line input displacement at some other point—by an actual process of topology/shape optimization.     

Power system model validation for power quality assessment applications using genetic algorithm

This paper presents an intelligent system for power quality assessment application. This system is used for power system model validation. A genetic algorithm (GA) based system for validating the power system model in capacitor switching studies has been developed. The problem formulation and the proposed solution are illustrated. The feasibility of the developed system for practical applications is demonstrated by evaluation studies.     

Optimized scenario for rainfall forecasting using genetic algorithm coupled with artificial neural network

Rainfall forecasting plays many important role in water resources studies such as river training works and design of flood warning systems. Recent advancement in artificial intelligence and in particular techniques aimed at converting input to output for highly nonlinear, non-convex and dimensionalized processes such as rainfall field, provide an alternative approach for developing rainfall forecasting model. Artificial neural networks (ANNs), which perform a nonlinear mapping between inputs and outputs, are such a technique. Current literatures on artificial neural networks show that the selection of network architecture and its efficient training procedure are major obstacles for their daily usage. In this paper, feed-forward type networks will be developed to simulate the rainfall field and a so-called back propagation (BP) algorithm coupled with genetic algorithm (GA) will be used to train and optimize the networks. The technique will be implemented to forecast rainfall for a number of times using rainfall hyetograph of recording rain gauges in the Upper Parramatta catchment in the western suburbs of Sydney, Australia. Results of the study showed the structuring of ANN network with the input parameter selection, when coupled with GA, performed better compared to similar work of using ANN alone.     

Information filtering using a probabilistic model

We propose an information filtering system based on a probabilistic model. We make an assumption that a document consists of words which occur according to a probability distribution, and regard a document as a sample drawn according to that distribution. In this article, we adopt a multinomial distribution and represent a document as probability which has random values as the words in the document. When an information filtering system selects information, it uses the similarity between the user's interests (a user profile) and a document. Since our proposed system is constructed under the probabilistic model, the similarity is defined using the Kullback Leibler divergence. To create the user profile, we must optimize the Kullback Leibler divergence. Since the Kullback Leibler divergence is a nonlinear function, we use a genetic algorithm to optimize it. We carry out experiments and confirm effectiveness of the proposed method. This work was presented in part at the 10th International Symposium on Artificial Life and Robotics, Oita, Japan, February 4–6, 2005     

Optimising a complex discrete event simulation model using a genetic algorithm

A steelworks model is selected as representative of the stochastic and unpredictable behaviour of a complex discrete event simulation model. The steel-works has a number of different entity or object types. Using the number of each entity type as parameters, it is possible to find better and worse combinations of parameters for various management objectives. A simple real-coded genetic algorithm is presented that optimises the parameters, demonstrating the versatility that genetic algorithms offer in solving hard inverse problems.     

Minimizing makespan on a three-machine flowshop batch scheduling problem with transportation using genetic algorithm

This paper studies the minimization of makespan in a three-machine flowshop scheduling problem in which a batch processing machine is located between two single processing machines on first and third stages. In this study also transportation capacity and transportation among machines times are explicitly considered.We establish a mixed integer programming model and propose a heuristic algorithm based on the basic idea of Johnson's algorithm. Since the problem under study is NP-hard, a genetic algorithm is also proposed to minimize makespan. The effectiveness of our solution procedures is evaluated through computational experiments. The results obtained from the computational study have shown that the genetic algorithm is a viable and effective approach that is capable to produce consistently good results.     

Efficient parallel algorithm for the two-dimensional diffusion equation subject to specification of mass

An efficient L ₀-stable parallel algorithm is developed for the two-dimensional diffusion equation with non-local time-dependent boundary conditions. The algorithm is based on subdiagonal Padé approximation to the matrix exponentials arising from the use of the method of lines and may be implemented on a parallel architecture using two processors running concurrently with each processor employing the use of tridiagonal solvers at every time-step. The algorithm is tested on two model problems from the literature for which discontinuities between initial and boundary conditions exist. The CPU times together with the associated error estimates are compared.