In 2 ¡Ñ2 case,we discuss the uniqueness of the u-invariant measure on projective space.Under the condition that |detM|=1 for any M in Gu and Gu is not compact,we have the followings: (1) For any x in P(R^2),if #{M¡Dx|M belongs Gu}>2, then the u-invariant measure is unique. (2) For some x in P(R^2),there exists x1,x2 such that {M¡Dx|M belongs Gu} is contained in {x1,x2},if x1 and x2 are both fixed,then the u-invariant measure v is not unique;otherwise,if u has mass only on x1 and x2,then the u-invariant measure is unique.

Let $mathcal{B}$ be a separable Banach space. In this thesis, it is shown that the Chung's strong law of large numbers holds for a sequence of independent $mathcal{B}$-valued random elements and an array of rowwise independent $mathcal{B}$-valued random elements under some weaker assumptions by using more generalized functions $phi_{n}$'s.

The first topic of this thesis is a voltage boost level clamping circuit for a flash memory which utilizes an implicit feedback loop as well as MOS transistors with different threshold voltages. The proposed design can be added to charge pumps to stabilize the output voltage. The unwanted output voltage spikes introduced by the linear pumping ratio are prevented. Not only are possible damages to memory cores avoided, the power disspation is reduced in contrast with prior regulator methods. The second topic is a switch-current 3-bit CMOS wideband random signal generator, which utilizes a digital normalizer to flatten the distribution of the probability in the entire range of B parameter. The ¡§colored¡¨ random numbers problem in prior designs is resolved. In addition, the coefficients of the proposed design are dynamically adjustable.

Abstract With the economic growth as well as the advent of the post-industry age in the 21st century, a variety of investment patterns for people to employ progressively appear. Moreover, there is a tendency in which people¡¦s traditional and conservative manners for investments gradually transform into aggressive ones with a preference of risks. For instance, it has been impossible nowadays for certain simple finance-managing mechanisms, such as the Certificate of Deposit, the Bond or even Stock Markets, to satisfy general investors¡¦ needs. Therefore, more and more financial products have become available, especially a lot of derivative securities like options, futures, or stock options. Furthermore, the activities of horseracing, dog racing or even lottery are no longer taboos in Taiwan. As far as the lotteries in most of the counties are concerned, the government is usually the main host, regarding them as additional effortless sources of tax revenue. During the earlier periods in Taiwan, the only lottery available was the national ¡§Ai-Guo¡¨ lottery issued by the government. Unfortunately, a lot of unlawful gambling gangs took advantage of the popularity of ¡§Ai-Guo¡¨ lottery by utilizing it as the sources of the so-called ¡§Da-Jia-Le¡¨ lottery, which was an illegal, yet popular underground gambling activity. The ¡§Da-Jia-Le¡¨ lottery not only corrupted the social values and caused a lot of criminal actions, but also made an acute and serious impact on regular economic operations. Consequently, the government resolutely called a halt on the ¡§Ai-Guo¡¨ lottery. Nevertheless, the prevailing gambling mania did decline; on the contrary, it further transformed into a covert and under-the-counter gambling operation attached to the Hong Kong lottery. The craze almost swept the entire country, resulting in more and more social problems and having troubled the authorities for a long period of time. To settle such a predicament, government decided to refer to the experiences of other countries, beginning to launch local lotteries in Taiwan, instead of prohibiting people from attending illicit gambling activities. On the one hand, the government would be able to set the lottery business back on right track and eliminate the root of the underground economic operations, which brought about many social problems. On the other hand, the local lottery would provide extra tax revenue for the administrations as well as more employment opportunities for handicapped people. Accordingly, Taiwanese Lottery was launched in the beginning of the year 2002. Until now, mainly due to the attraction of the surprisingly huge bonus shares offered by the lottery¡¦s highest prize, the public response is overwhelming. The social environment in which the stochastic investment actions become authorized and gradually prevail directly triggers the motivation of this research in further investigating the feasibility of the investment. The research takes Taiwan Public Welfare Lottery as an investment research case, exploring the game rules of lottery in an attempt to enhance the probability of winning prizes and further testing and evaluating the effectiveness of various simple biting tools adopted by the general public in order to obtain the most effective biting ways for individual investor¡¦s reference. The process of this study is as follows: 1. Testing the randomness of the winning prize numbers in Taiwan Public Welfare Lottery. 2. Estimating the possible distribution of the population via the statistical application. 3. Using the cluster analysis to induce the sub-cluster of the selected numbers according to the patterns of the historical prize numbers. 4. Making use of Markov Chain to select the sub-cluster, which is within an investor¡¦s upper limit of risks. And consequently, this study reveals the followings: 1. The winning prize numbers are random within study samples. 2. The statistic order of the winning prize numbers is £] distribution. 3. The cluster effect is slightly enhanced when the sub-cluster is chosen through the cluster analysis rather than random ways. 4. The application of Markov Chain has an obvious reinforcement on the selection of sub-cluster. 5. The so-called ¡§Smart Package Biting¡¨, which is adopted by the general public, cannot improve the winning possibility; however, depending on individual investor¡¦s needs, it could be a useful tool for risk management. Through a seemingly aimless random event, the research explores the possible development of the modeling process, so as to comprehend and experience a manager¡¦s mental course on decision making in a dynamic, capricious or even chaotic managerial environment.

Over the last few years, peer-to-peer (P2P) applications have relentlessly grown to represent a formidable component of Internet traffic. In contract to P2P networks witch used well-defined port number, current P2P applications have use of arbitrary ports. As P2P applications continue to evolve, robust and effective methods are methods are needed for P2P traffic identification. Many P2P applications are bandwidth-intensive. Understanding the Internet traffic profile is important for several reasons, including traffic engineering, network service pricing. In this Thesis, we integrated port-based method into original Classifier which is using content-based method only. Therefore, we can improve the recognition rate for Classifier and identify more applications. We also verified our Classifier recognition rate by using the results of Service Control Engine.

A 2-D fully nonlinear numerical wave tank (NWT) is developed to investigate the Bragg resonance scattered by submerged multi-array breakwaters for random waves. This model is based on a boundary integral equation method with linear element scheme. The fully nonlinear free surface boundary condition is treated using the Mixed Eulerian-Lagrangian method and the 4th-order Runge-Kutta method. The incident random waves are generated by JONSWAP spectrum at one end of the wave tank. Two damping zones are deployed at both ends of the NWT to absorb the energy of the reflected and transmitted waves. In the regular wave cases, the results of Bragg reflection calculated are in good agreement with that of other experiments and numerical models. In addition, the simulated spectrum of random waves is also verified by the original input spectrum. The results of the random waves have the same trend as those of the Regular waves. The reflection coefficient for random waves at the first peak of resonance is about 70 percent of that of the regular wave, but the frequency of band width of Bragg effect has become wider and this advantage may compensate the peak reduction. Finally, we may conclude that the present model is adequate to use as a tool for coastal protection. Systematic studies for random waves propagating over series submerged breakwaters are conducted. The Bragg reflection will be enhanced with the increase of relative height, the length of bars, the number of breakwaters, and the toe angle of submerged breakwaters. In this study, it also reveals that the frequency of peak reflection for higher breakwaters has down shift phenomenon.

none

The desire is infinite, but the resources are limited. The relationship between fiscal decentralization and long-run economic growth is ambiguous. Several economists have made the case for fiscal decentralization as a means of promoting long-run economic growth based on the view that it leads to better resource allocation and a more productive, and possibly smaller, public sector. Countries have pursued decentralization policies both for political and developmental reasons. Fiscal decentralization, the allocation of tax and spending powers to lower levels of government, currently in vogue is based on notions of economic efficiency criteria. Although it is theoretically expected that decentralization leads to efficient provision of local public services and stimulates economic development, the theoretical underpinnings for this relationship remain largely undeveloped. The absence of an adequate theoretical framework has undermined the validity of the empirical work on this subject. Advocates of fiscal decentralization argue that among other benefits, it can increase the efficiency of delivery of government services. We use data from 23 counties (cities) of the Taiwan province. The empirical findings can be stated as follows. The primary finding is that the estimated coefficient on fiscal decentralization variable is positive and statistically significant . This finding provides evidence that fiscal decentralization contributes to economic growth. This paper is also one of the first to evaluate this claim empirically by looking at the association between expenditure decentralization and the efficiency of local government .We also provide evidence that expenditure (revenue) decentralization is a negative (positive) relation with the efficiency of local government. Further Tobit panel regression of 23 counties (cities) provide robust evidence that more decentralization is not associated with higher efficiency of local governments. Therefore that fiscal decentralization contributes to the efficiency of local governments is ambiguous in the previous period.

In Chinese language, the smallest meaningful unit is a word which is composed of a sequence of characters. A Chinese sentence is composed of a sequence of words without any separation between them. In the area of information retrieval or data mining, the segmentation of a sequence of Chinese characters should be done before anyone starts to use these segments of characters. The process is called the Chinese word segmentation. The researches of Chinese word segmentation have been developed for many years. Although some recent researches have achieved very high performance, the recall of those words that are not in the dictionary only achieves sixty or seventy percent. An approach described in this paper makes use of the linear-chain conditional random fields (CRFs) to have a more accurate Chinese word segmentation. The discriminatively trained model that uses two of our proposed feature templates for deciding the boundaries between characters is used in our study. We also propose three other methods, which are the duplicate word repartition, the date representation repartition, and the segment refinement, to enhance the accuracy of the processed segments. In the experiments, we use several different approaches for testing and compare the results with those proposed by Li et al. and Lau and King based on three different Chinese word corpora. The results prove that the improved feature template which makes use of the information of prefix and postfix could increase both the recall and the precision. For example, the F-measure reaches 0.964 in the MSR dataset. By detecting repeat characters, the duplicated characters could also be better repartitioned without using extra resources. In the representation of date, the wrongly segmented date could be better repartitioned by using the proposed method which deals with numbers, date, and measure words. If a word is segmented differently from that of the corresponding standard segmentation corpus, a proper segment could be produced by repartitioning the assembled segment which is composed of the current segment and the adjacent segment. In the area of using the conditional random fields for Chinese word segmentation, we have proposed a feature template for better result and three methods which focus on other specific segmentation problems.

<p>Analyzing and modeling data is a well established research area and a vast variety of different methods have been developed over the last decades. Most of these methods assume fixed positions of data points; only recently uncertainty in data has caught attention as potentially useful source of information. In order to provide a deeper insight into this subject, this thesis concerns itself with the following essential question: Can information on uncertainty of feature values be exploited to improve in silico modeling? For this reason a state-of-art random forest algorithm is developed using Matlab R. In addition, three techniques of handling uncertain numeric features are presented and incorporated in different modified versions of random forests. To test the hypothesis six realworld data sets were provided by AstraZeneca. The data describe biochemical features of chemical compounds, including the results of an Ames test; a widely used technique to determine the mutagenicity of chemical substances. Each of the datasets contains a single uncertain numeric feature, represented as an expected value and an error estimate. Themodified algorithms are then applied on the six data sets in order to obtain classifiers, able to predict the outcome of an Ames test. The hypothesis is tested using a paired t-test and the results reveal that information on uncertainty can indeed improve the performance of in silico models.</p>