Recent years have witnessed increasing interest in using metaheuristics to solve the codebook generation problem (CGP) of vector quantization as well as increasing interest in reducing the computation time of metaheuristics. One of the recently proposed methods aimed at reducing the computation time of metaheuristics is based on the notion of pattern reduction (PR). The problem with PR is in that it may compress and remove patterns that are not supposed to be compressed and removed, thus decreasing the quality of the solution. In this thesis, we proposed a fuzzy version of PR called fuzzy pattern reduction (FPR) to reduce the possibility of compressing and removing patterns that are not supposed to be compressed and removed. To evaluate the performance of the proposed algorithm, we apply it to the following four metaheuristics: generalized Lloyd algorithm, code displacement, genetic k-means algorithm, and particle swarm optimization and use them to solve the CGP. Our experimental results show that the proposed algorithm can not only significantly reduce the computation time but also improve the quality of all the metaheuristics evaluated.

The photodissociation dynamics of cyanogen iodide (ICN) in the A continuum was studied by the three-dimensional sliced fluorescence imaging (3DSFI) method. The fourth harmonic of a Nd:YAG laser (266 nm) was used to dissociate a low pressure (5 mTorr) sample of ICN, and the resulting CN photofragments were probed by optical-optical double resonance (OODR) detection scheme. A theoretical framework on the emitted fluorescence intensity has been developed to analyze the experimentally obtained image. The vector correlations among the ICN parent molecule transition dipole moment £g, the CN fragment recoil velocity v and the CN rotational angular moment J can be revealed.

Given a protein sequence and the C£\ coordinates on its backbone, the all-atom protein backbone reconstruction problem (PBRP) is to reconstruct the backbone by its 3D coordinates of N, C and O atoms. In the past few decades, many methods have been proposed for solving PBRP, such as ab initio, homology modeling, SABBAC, Wang¡¦s method, Chang¡¦s method, BBQ (Backbone Building from Quadrilaterals) and Chen¡¦s method. Chen found that, if they can choose the correct prediction tool to build the 3D coordinates of the desired atoms, the RMSD may be improved. In this thesis, we propose a method for solving PBRP based on Chen¡¦s method. We use tool preference classification on each atom of the residue, where the classification model is generated by SVM (Support Vector Machine). We rebuild the backbone by combing the prediction results of all atoms in all residues. The data sets used in our experiments are CASP7, CASP8 and CASP9, which have 65, 52 and 63 proteins, respectively. These data sets contain nonstandard amino acids as well as standard ones. We improve the average RMSDs of Chen¡¦s results in some cases. The average RMSDs of our method are 0.3496 in CASP7, 0.3084 in CASP8 and 0.3286 in CASP9.

In this paper, we propose a novel approach of vector quantization using a merge-based hierarchical neural network. Vector quantization¡]VQ¡^is known as a very useful technique for lossy data compression. Recently, Neural network¡]NN¡^algorithms have been used for VQ. Vlajic and Card proposed a modified adaptive resonance theory (modified ART2¡^[1] which is a constructing tree structure clustering method. However, modified ART2 has disadvantages of slow construction rate and constructing many redundant levels. Therefore, we propose a more efficient approach for constructing the tree in this paper. Our method establishes only those required levels without losing the fidelity of a compressed image.

To resolve the inflation caused by overheated economy in 1992, China executed Macroscopic Control Policy to stabilize the fluctuation of price standard in 1993. It seemed to achieve the effort of controlling inflation. However, because of longtime Macroscopic Control Policy after Asian Financial Crisis, it resulted in negative impacts. CPI in China has been minus quantity for 39 months from October in 1997 to December in 2000. And CPI turned plus into minus from April in 1998 to January in 2000. And CPI turned plus into minus from April in 1998 to January in 2000. The growth rate of RPI is -2.6% and that of CPI is -0.8% in 1998. It declined to -3.0%(RPI) and -1.4%(CPI) in 1999. The growth rate of GDP has fallen down since 1992. The main purpose of this paper is to explore the reason of the deflation late in 1990 in China. I want to find out why deflation was happened in china? What is the main cause of deflation in china? What are the impacts and shocks to china economic growth by these causes? How are the impacts and shocks to china economic growth by these causes? The whole supply and demand and money contraction resulted in the downfall of GDP and CPI. To prevent the phenomena of overheated economy since 1993, most of investment moved away China because of Macroscopic Control Policy. Under this kind of situation, we couldn¡¦t say that the investments were excess. Therefore, the main reason isn¡¦t prices dropping caused by too much supply. China continued Deflation Policy after Asian Financial Crisis in 1997, so the speed of economy development decreased slowly. It also reduced the whole consumption, public spending, investment export, and money supply and demand. The effect of negative development resulted in the deflation of economy development.

The problems of the scattering off the perfect conductor sphere and the dielectric sphere have the exact solutions. But there are no exact solutions for the scattering off a multi-layered dielectric sphere with a metallic cap like the Lunberg lens reflectors which is used as a strong omni-directional reflector found in many microwave applications. Haruo Sakurai applied the modal expansion technique and point-matching method (PMM) to study the scattering of the Lunberg lens reflectors. The problem is eventually formulated as 2MN by 2MN simultaneous matrix equation with M regions each having 2N unknowns due to two set of coupled polarization vectors. Strictly speaking, the formulae of the mode matching method for the problem of the scattering of the dielectric sphere are not exact. Furthermore, the size of the simultaneous matrix equation is also unnecessarily too larger. In this thesis, we employ an integral equation formulation in the Frequency-domain together with the modified impedance transformation technique for the spherically layers to study the scattering of the Lunberg lens reflectors. We show that the formulae of the integral equation are exact and using an equivalent matrix equation, that the entire problem can be reduced to a N by N matrix equation where N is the number of terms of the expansion of the unknown field in the opening. To verify our formulation we compute the total field of the plane wave incident upon the multi-layered micro lenses and compared the results with those from the geometric optics. We get good agreement for the regions that both theories apply. Small discrepancy is also observed and is consistent with the theoretical prediction.

In response to the global climate change and environmental protection needs, more and more nations take renewable energies as one of the major future energy policies for its characteristics of clean, low greenhouse gas emission and self-productivity. Among Taiwan¡¦s many renewable energy development, wind energy is always on the list. Wind energy will not only avoid the problems of carbon dioxide which cause greenhouse effect, but also refrain pollution caused by fossil fuel or nuclear energy. The induction generator is a fundamental component of a wind machine. It¡¦s electric port and control can be classified into single fed control and double fed control. Through various comparisons of control and operation, double fed control seems to be more stable at present. Double fed induction generator may continuously produce power at island state without the supply of reactive power from parallel capacitor or the synchronous generator. which is the biggest advantage than single fed induction. , In other words, double fed induction generator may supply power independently. The induction generator is a time-varying and non-linear component. So the traditional rule of scalar control is not adopted in the thesis. The induction generator can be linearized by using vector control, and the power flow can also be directed with the control of its composition of rotor current vector and the controller is designed under the above methodology. The thesis aims to simulate, analyze and compare the steady and transient states of single fed and double fed wind generators with the purposes a more through understanding of the controller and its operation.

During the Digital Period, the more people using these digital video. When there are more and more users and amount of video data, the management of video data becomes a significant dimension during development. Therefore, there are more and more studying of accomplishing video database system, which provide users to search and get them. In this paper, a novel method for Video Scene Change Detection and video database retrieval is proposed. Uses Fractal orthonormal bases to guarantee the similar index has the similar image the characteristic union support vector clustering, splits a video into a sequence of shots, extracts a few representative frames(key-frames) to take the video database index from each shot. When image search compared to, according to MIL to pick up the characteristic, which images pursues the video database to have the similar characteristic, computation similar, makes the place output according to this.

There are three purposes in this thesis. First of all, the thesis examines the market herding behavior of real estates in Taiwan. Second, if herding exists in the markets, what factors cause herding behavior? Finally, how does herding behavior affect market performance? The empirical results show there is no herding evidence in the real estates markets when market moves downward. There are different results between different real estates and periods when market moves upward. There is obvious herding behavior in the pre-sale houses in Taipei City. Herding behavior also exist lag phenomenon. During lag period, there is obvious herding behavior in the pre-sale houses in Taipei when market moves upward. There is weak herding behavior in the other real estates, including pre-sale houses of Taipei County, existing houses and land. We find that macro factors such as money supply, GDP growth rate, saving amount, incoming and micro factors such as market depth and market return in the last period will cause herding behavior. Finally, herding behavior, price volatility and transaction amount volatility have Granger causality. Herding behavior will makes price more efficient and increases transaction.

The central government and most of local governments in Taiwan have adopted the e-mail services to provide citizens for requesting services or expressing their opinions through Internet. Traditionally, these requests/opinions need to be manually classified into appropriate departments for service rendering. However, due to the ever-increasing number of requests/opinions received, the manual classification approach is time consuming and becomes impractical. Therefore, in this study, we attempt to apply text categorization techniques for constructing automatically a classification mechanism in order to establish an efficient e-government service portal. The purpose of this thesis is to investigate effectiveness of different text categorization methods in supporting automatic classification of service requests/opinions emails sent to Mayor¡¦s mailbox. Specifically, in each phase of text categorization learning, we adopt and evaluate two methods commonly employed in prior research. In the feature selection phase, both the maximal x2¡@statistic method and the weighted average x2¡@statistic method of x2¡@statistic are evaluated. We consider the Binary and TFxIDF representation schemes in the document representation phase. Finally, we adopt the decision tree induction technique and the support vector machines (SVM) technique for inducing a text categorization model for our target e-government application. Our empirical evaluation results show that the text categorization method that employs the maximal x2 statistic method for feature selection, the Binary representation scheme, and the support vector machines as the underlying induction algorithm can reach an accuracy rate of 77.28% and an recall and precision rates of more than 77%. Such satisfactory classification effectiveness suggests that the text categorization approach can be employed to establish an effective and intelligent e-government service portal.