C*-algebras generalizing Cuntz-Krieger algebras can be associated to hyperbolic homeomorphisms of compact metric spaces. They satisfy a non-commutative form of Spanier-Whitehead duality with respect to K-theory. We prove this for the case of subshifts of finite type. The special feature of the present situation is that the constructions are all done on the full Fock space and are very explicit, while the general theorem requires much more abstract machinery.

Analyzing test particles falling into a Kerr black hole, we study gravitational waves in Brans-Dicke theory of gravity. First we consider a test particle plunging with a constant azimuthal angle into a rotating black hole and calculate the waveform and emitted energy of both scalar and tensor modes of gravitational radiation. We find that the waveform as well as the energy of the scalar gravitational waves weakly depends on the rotation parameter of black hole $a$ and on the azimuthal angle. Secondly, using a model of a non-spherical dust shell of test particles falling into a Kerr black hole, we study when the scalar modes dominate. When a black hole is rotating, the tensor modes do not vanish even for a ``spherically symmetric" shell, instead a slightly oblate shell minimizes their energy but with non-zero finite value, which depends on Kerr parameter $a$. As a result, we find that the scalar modes dominate only for highly spherical collapse, but they never exceed the tensor modes unless the Brans-Dicke parameter $\omega_{BD} \lsim 750 $ for $a/M=0.99$ or unless $\omega_{BD} \lsim 20,000 $ for $a/M=0.5$, where $M$ is mass of black hole. We conclude that the scalar gravitational waves with $\omega_{BD} \lsim$ several thousands do not dominate except for very limited situations (observation from the face-on direction of a test particle falling into a Schwarzschild black hole or highly spherical dust shell collapse into a Kerr black hole). Therefore observation of polarization is also required when we determine the theory of gravity by the observation of gravitational waves.

We examine the electronic states of Sr$_{14-x}$Ca$_x$Cu$_{24}$O$_{41}$ by the ionic and cluster model approach. It is found that self-doped holes are likely to stay on the chain at $x=0$ and the Ca substitution drives the holes to move to the ladder. This feature is caused by the change of the positions of (Sr,Ca) layers, which enhances the electrostatic potentials in the chain. The optical conductivity is calculated applying the exact diagonalization method to small Cu-O clusters. It is shown that the excitations below 2 eV are governed by the ladder, giving rise to the spectral weight transfer from the charge-transfer excitation around 2 eV to low energy Drude excitation, while the contribution from the chain mainly emerges in higher energy region showing large weight around 3 eV. These provide consistent explanation to recent experiment.

The Maximum Entropy Modeling Toolkit supports parameter estimation and prediction for statistical language models in the maximum entropy framework. The maximum entropy framework provides a constructive method for obtaining the unique conditional distribution p*(y|x) that satisfies a set of linear constraints and maximizes the conditional entropy H(p|f) with respect to the empirical distribution f(x). The maximum entropy distribution p*(y|x) also has a unique parametric representation in the class of exponential models, as m(y|x) = r(y|x)/Z(x) where the numerator m(y|x) = prod_i alpha_i^g_i(x,y) is a product of exponential weights, with alpha_i = exp(lambda_i), and the denominator Z(x) = sum_y r(y|x) is required to satisfy the axioms of probability. This manual explains how to build maximum entropy models for discrete domains with the Maximum Entropy Modeling Toolkit (MEMT). First we summarize the steps necessary to implement a language model using the toolkit. Next we discuss the executables provided by the toolkit and explain the file formats required by the toolkit. Finally, we review the maximum entropy framework and apply it to the problem of statistical language modeling. Keywords: statistical language models, maximum entropy, exponential models, improved iterative scaling, Markov models, triggers.

The claim put forward in [hep-th/9512051, hep-th/9612244] that the energies of the ``missing'' states of three anyons in a harmonic potential depend linearly on the statistics parameter, is incorrect because the wave functions proposed do not satisfy the anyonic interchange conditions.

We have measured the distribution function of the flux decrement caused by Lyman alpha forest absorption in a new sample of high resolution QSO spectra. The observations are compared to the results from two simulations of the Lya forest: an Eulerian Lambda-CDM model, and an SPH standard CDM model. Good agreement between the shapes of simulated and observed distributions is achieved by globally scaling the optical depth to match the mean flux decrements. This procedure amounts to a measurement of the parameter Omega_b^2 h^3 / Gamma (where Omega_b is the baryonic matter density and Gamma is the HI ionization rate). Estimating a lower limit Gamma > 7 10^{-13} s^{-1} from the abundance of known QSOs, we derive a lower limit to the baryon density, Omega_b h^2>0.021(0.017) for the Lambda-CDM (SCDM) model. In both cases the large values are inconsistent with some recent D/H determinations (Rugers & Hogan 1996a,b), favoring a low deuterium abundance as reported by Tytler, Fan & Burles (1996). Adopting a fixed Omega_b, we can determine the evolution of the ionizing radiation field. Our models predict the intensity to be approximately constant with redshift, consistent with the assumption that the ionizing background is produced by known quasars for z < 3. However, additional sources of ionizing photons are required at higher redshift.

Instance-based learning techniques typically handle continuous and linear input values well, but often do not handle nominal input attributes appropriately. The Value Difference Metric (VDM) was designed to find reasonable distance values between nominal attribute values, but it largely ignores continuous attributes, requiring discretization to map continuous values into nominal values. This paper proposes three new heterogeneous distance functions, called the Heterogeneous Value Difference Metric (HVDM), the Interpolated Value Difference Metric (IVDM), and the Windowed Value Difference Metric (WVDM). These new distance functions are designed to handle applications with nominal attributes, continuous attributes, or both. In experiments on 48 applications the new distance metrics achieve higher classification accuracy on average than three previous distance functions on those datasets that have both nominal and continuous attributes.

The phase diagram of quantum electron bilayers in zero magnetic field is obtained using density functional theory. For large electron densities the system is in the liquid phase, while for smaller densities the liquid may freeze (Wigner crystallization) into four different crystalline phases; the lattice symmetry and the critical density depend on the the inter-layer distance. The phase boundaries between different Wigner crystals consist of both first and second order transitions, depending on the phases involved, and join the freezing curve at three different triple points.

Previous approaches of analyzing spontaneously spoken language often have been based on encoding syntactic and semantic knowledge manually and symbolically. While there has been some progress using statistical or connectionist language models, many current spoken- language systems still use a relatively brittle, hand-coded symbolic grammar or symbolic semantic component. In contrast, we describe a so-called screening approach for learning robust processing of spontaneously spoken language. A screening approach is a flat analysis which uses shallow sequences of category representations for analyzing an utterance at various syntactic, semantic and dialog levels. Rather than using a deeply structured symbolic analysis, we use a flat connectionist analysis. This screening approach aims at supporting speech and language processing by using (1) data-driven learning and (2) robustness of connectionist networks. In order to test this approach, we have developed the SCREEN system which is based on this new robust, learned and flat analysis. In this paper, we focus on a detailed description of SCREEN's architecture, the flat syntactic and semantic analysis, the interaction with a speech recognizer, and a detailed evaluation analysis of the robustness under the influence of noisy or incomplete input. The main result of this paper is that flat representations allow more robust processing of spontaneous spoken language than deeply structured representations. In particular, we show how the fault-tolerance and learning capability of connectionist networks can support a flat analysis for providing more robust spoken-language processing within an overall hybrid symbolic/connectionist framework.

We discuss an extention of the Minimal Supersymmetric Standard Model (MSSM) with the 4th and anti-4th generations which have $SU(2)_L\times U(1)_Y$ invariant masses. Due to the the extra generations, all three running gauge couplings become asymptotically non--free while preserving gauge coupling unification at the GUT scale. We show that due to the asymptotically non--free character of the gauge couplings: (1) the top and bottom Yukawa couplings are strongly focused onto infrared fixed points as they are evolved down in scale making their values at $\mu=\mz$ insensitive to their initial values at $\mu=\mgut$; (2) the model predicts $\R(\mz) \equiv Y_b/Y_\tau |_{\mu =\mz}\approx 1.8$, which is consistent with the experimental value provided we take the ratio of Yukawa couplings at the GUT scale to be $\R(\mgut) = Y_b/Y_\tau |_{\mu = \mgut} = 1/3$; (3) the $t$ mass prediction comes out to be $\mt\approx 180\,\GeV$ which is also consistent with experiment.