Improved Nonlinear PCA for Process Monitoring Using Support Vector Data Description

Nonlinear principal component analysis (PCA) based on neural networks has drawn significant attention as a monitoring tool for complex nonlinear processes, but there remains a difficulty with determining the optimal network topology. This paper exploits the advantages of the Fast Recursive Algorithm, where the number of nodes, the location of centres, and the weights between the hidden layer and the output layer can be identified simultaneously for the radial basis function (RBF) networks. The topology problem for the nonlinear PCA based on neural networks can thus be solved. Another problem with nonlinear PCA is that the derived nonlinear scores may not be statistically independent or follow a simple parametric distribution. This hinders its applications in process monitoring since the simplicity of applying predetermined probability distribution functions is lost. This paper proposes the use of a support vector data description and shows that transforming the nonlinear principal components into a feature space allows a simple statistical inference. Results from both simulated and industrial data confirm the efficacy of the proposed method for solving nonlinear principal component problems, compared with linear PCA and kernel PCA.

A splitting result for the algebraic K-theory of projective toric schemes

Suppose X is a projective toric scheme defined over a ring R and equipped with an ample line bundle L . We prove that its K-theory has a direct summand of the form K(R)(k+1) where k = 0 is minimal such that L?(-k-1) is not acyclic. Using a combinatorial description of quasi-coherent sheaves we interpret and prove this result for a ring R which is either commutative, or else left noetherian.

Scaling of Food-Web Properties with Diversity and Complexity Across Ecosystems

Trophic scaling models describe how topological food-web properties such as the number of predator prey links scale with species richness of the community. Early models predicted that either the link density (i.e. the number of links per species) or the connectance (i.e. the linkage probability between any pair of species) is constant across communities. More recent analyses, however, suggest that both these scaling models have to be rejected, and we discuss several hypotheses that aim to explain the scale dependence of these complexity parameters. Based on a recent, highly resolved food-web compilation, we analysed the scaling behaviour of 16 topological parameters and found significant power law scaling relationships with diversity (i.e. species richness) and complexity (i.e. connectance) for most of them. These results illustrate the lack of universal constants in food-web ecology as a function of diversity or complexity. Nonetheless, our power law scaling relationships suggest that fundamental processes determine food-web topology, and subsequent analyses demonstrated that ecosystem-specific differences in these relationships were of minor importance. As such, these newly described scaling relationships provide robust and testable cornerstones for future structural food-web models.

Transmission-Line Class-E Power Amplifier With Extended Maximum Operating Frequency

A novel Class-E power amplifier (PA) topology with transmission-line load network is presented in this brief. When compared with the classic Class-E topology, the new circuit can increase the maximum operating frequency up to 50% higher without trading the other Class-E figures of merit. Neither quarterwave line/massive radio-frequency choke for collector/drain biasing nor additional fundamental-frequency output matching circuit are needed in the proposed PA, thus resulting in a compact design. Closed-form formulations are derived and verified by simulations with practical design limitations carefully taken into consideration and good agreement achieved.

Finite domination and Novikov rings. Iterative approach

Suppose C is a bounded chain complex of finitely generated free modules over the Laurent polynomial ring L = R[x,x -1]. Then C is R-finitely dominated, i.e. homotopy equivalent over R to a bounded chain complex of finitely generated projective R-modules if and only if the two chain complexes C ? L R((x)) and C ? L R((x -1)) are acyclic, as has been proved by Ranicki (A. Ranicki, Finite domination and Novikov rings, Topology 34(3) (1995), 619–632). Here R((x)) = R[[x]][x -1] and R((x -1)) = R[[x -1]][x] are rings of the formal Laurent series, also known as Novikov rings. In this paper, we prove a generalisation of this criterion which allows us to detect finite domination of bounded below chain complexes of projective modules over Laurent rings in several indeterminates.

Inline Interdigital Pseudo-Elliptic Helical Resonator Filters

A new inline coupling topology for narrowband helical resonator filters is proposed that allows to introduce selectively located transmission zeros (TZs) in the stopband. We show that a pair of helical resonators arranged in an interdigital configuration can realize a large range of in-band coupling coefficient values and also selectively position a TZ in the stopband. The proposed technique dispenses the need for auxiliary elements, so that the size, complexity, power handling and insertion loss of the filter are not compromised. A second order prototype filter with dimensions of the order of 0.05 lambda, power handling capability up to 90 W, measured insertion loss of 0.18 dB and improved selectivity is presented.

Gaussian Approximation for Tracking Occluding and Interacting Targets

In this paper, we show how interacting and occluding targets can be tackled successfully within a Gaussian approximation. For that purpose, we develop a general expansion of the mean and covariance of the posterior and we consider a first order approximation of it. The proposed method differs from EKF in that neither a non-linear dynamical model nor a non-linear measurement vector to state relation have to be defined, so it works with any kind of interaction potential and likelihood. The approach has been tested on three sequences (10400, 2500, and 400 frames each one). The results show that our approach helps to reduce the number of failures without increasing too much the computation time with respect to methods that do not take into account target interactions.

Allometric analysis of Sitka spruce branches:Mechanical versus hydraulic design principles

The geometry of tree branches can have considerable effect on their efficiency in terms of carbon export per unit carbon investment in structure. The purpose of this study was to evaluate different design criteria using data describing the form of Picea sitchensis branches. Allometric analysis of the data suggests that resources are distributed to favour shoots with the greatest opportunity for extension into new space, with priority to the extension of the leader. The distribution of allometric relations of links (branch elements) was tested against two models: the pipe model, based on hydraulic transport requirements, and a static load model based on the requirement of shoots to provide mechanical resistance to static loads. Static load resistance required the load parameter to be proportional to the link radius raised to the power of 4. This was shown to be true within a 95% statistical confidence limit. The pipe model would require total distal length to be proportional to link radius squared but the measured branches did not conform well to this model. The comparison suggests that the diameters of branch elements were more related to the requirements for mechanical load. The cost of following a hydraulic design principle (the pipe model) in terms of mechanical efficiency was estimated and suggested that the pipe model branch would not be mechanically compromised but would use structural resources inefficiently. Resource allocation among branch elements was found to be consistent with mechanical stability criteria but also indicated the possibility of allocation based on other criteria, such as potential light interception by shoots. The evidence suggests that whilst branch topology increments by reiteration of units of morphogenesis, the geometry follows a functional design pattern.

Disjoint mixing operators

Chan and Shapiro showed that each (non-trivial) translation operator acting on the Fréchet space of entire functions endowed with the topology of locally uniform convergence supports a universal function of exponential type zero. We show the existence of d-universal functions of exponential type zero for arbitrary finite tuples of pairwise distinct translation operators. We also show that every separable infinite-dimensional Fréchet space supports an arbitrarily large finite and commuting disjoint mixing collection of operators. When this space is a Banach space, it supports an arbitrarily large finite disjoint mixing collection of C0-semigroups. We also provide an easy proof of the result of Salas that every infinite-dimensional Banach space supports arbitrarily large tuples of dual d-hypercyclic operators, and construct an example of a mixing Hilbert space operator T so that (T,T2) is not d-mixing.

Structural and functional relationships in the virulence-associated cathepsin L proteases of the parasitic liver fluke, Fasciola hepatica

The helminth parasite Fasciola hepatica secretes cysteine proteases to facilitate tissue invasion, migration, and development within the mammalian host. The major proteases cathepsin L1 (FheCL1) and cathepsin L2 (FheCL2) were recombinantly produced and biochemically characterized. By using site-directed mutagenesis, we show that residues at position 67 and 205, which lie within the S2 pocket of the active site, are critical in determining the substrate and inhibitor specificity. FheCL1 exhibits a broader specificity and a higher substrate turnover rate compared with FheCL2. However, FheCL2 can efficiently cleave substrates with a Pro in the P2 position and degrade collagen within the triple helices at physiological pH, an activity that among cysteine proteases has only been reported for human cathepsin K. The 1.4-A three-dimensional structure of the FheCL1 was determined by x-ray crystallography, and the three-dimensional structure of FheCL2 was constructed via homology-based modeling. Analysis and comparison of these structures and our biochemical data with those of human cathepsins L and K provided an interpretation of the substrate-recognition mechanisms of these major parasite proteases. Furthermore, our studies suggest that a configuration involving residue 67 and the "gatekeeper" residues 157 and 158 situated at the entrance of the active site pocket create a topology that endows FheCL2 with its unusual collagenolytic activity. The emergence of a specialized collagenolytic function in Fasciola likely contributes to the success of this tissue-invasive parasite.

Conformational Dynamics of the Human Propeller Telomeric DNA Quadruplex on a Microsecond Time Scale

The human telomeric DNA sequence with four repeats can fold into a parallel-stranded propeller-type topology. NMR structures solved under molecular crowding experiments correlate with the crystal structures found with crystal-packing interactions that are effectively equivalent to molecular crowding. This topology has been used for rationalization of ligand design and occurs experimentally in a number of complexes with a diversity of ligands, at least in the crystalline state. While G-quartet stems have been well characterised, the interactions of the TTA loop with the G-quartets are much less defined. To better understand the conformational variability and structural dynamics of the propeller-type topology, we performed molecular dynamics simulations in explicit solvent up to 1.5 µs. The analysis provides a detailed atomistic account of the dynamic nature of the TTA loops highlighting their interactions with the G-quartets including formation of an A:A base pair, triad, pentad and hexad. The results present a threshold in quadruplex simulations, with regards to understanding the flexible nature of the sugar-phosphate backbone in formation of unusual architecture within the topology. Furthermore, this study stresses the importance of simulation time in sampling conformational space for this topology.

On Quillen's calculation of graded K-theory

We adapt Quillen’s calculation of graded K-groups of Z-graded rings with support in N to graded K-theory, allowing gradings in a product Z×G with G an arbitrary group. This in turn allows us to use induction and calculate graded K-theory of Z -multigraded rings.

Ultrafast low-loss 42-70 GHz differential SPDT switch in 0.35 μm SiGe technology

This paper presents an ultrafast wideband low-loss single-pole double-throw (SPDT) differential switch in 0.35 µ m SiGe bipolar technology. The proposed topology adopting current-steering technique results in a total measured switching time of 75 ps , which suggests a maximum switching rate of 13 Gb/s, the fastest ever reported at V-band. In addition, the switch exhibits an insertion loss lower than 1.25 dB and an isolation higher than 18 dB from 42 GHz to 70 GHz. © 2006 IEEE.

A multiagent network for peer norm enforcement

In a multiagent system where norms are used to regulate the actions agents ought to execute, some agents may decide not to abide by the norms if this can benefit them. Norm enforcement mechanisms are designed to counteract these benefits and thus the motives for not abiding by the norms. In this work we propose a distributed mechanism through which agents in the multiagent system that do not abide by the norms can be ostracised by their peers. An ostracised agent cannot interact anymore and looses all benefits from future interactions. We describe a model for multiagent systems structured as networks of agents, and a behavioural model for the agents in such systems. Furthermore, we provide analytical results which show that there exists an upper bound to the number of potential norm violations when all the agents exhibit certain behaviours. We also provide experimental results showing that both stricter enforcement behaviours and larger percentage of agents exhibiting these behaviours reduce the number of norm violations, and that the network topology influences the number of norm violations. These experiments have been executed under varying scenarios with different values for the number of agents, percentage of enforcers, percentage of violators, network topology, and agent behaviours. Finally, we give examples of applications where the enforcement techniques we provide could be used.

Design technique for mm-wave IC realization of the load network of switched-mode class-EF power amplifier

Analysis and synthesis of the new Class-EF power amplifier (PA) are presented in this paper. The proposed circuit offers means to alleviate some of the major issues faced by existing Class-EF and Class-EF PAs, such as (1) substantial power losses due to parasitic resistance of the large inductor in the Class-EF load network, (2) unpredictable behaviour of practical lumped inductors and capacitors at harmonic frequencies, and (3) deviation from ideal Class-EF operation mode due to detrimental effects of device output inductance at high frequencies. The transmission-line load network of the Class-EF PA topology elaborated in this paper simultaneously satisfies the Class-EF optimum impedance requirements at fundamental frequency, second, and third harmonics as well as simultaneously providing matching to the circuit optimum load resistance for any prescribed system load resistance. Furthermore, an elegant solution using an open and short-circuit stub arrangement is suggested to overcome the problem encountered in the mm-wave IC realizations of the Class-EF PA load network due to lossy quarter-wave line. © 2010 IEICE Institute of Electronics Informati.