Food web of macroinvertebrate community in a Yangtze shallow lake: trophic basis and pathways

No detailed food web research on macroinvertebrate community of lacustrine ecosystem was reported in China. The present study is the first attempt on the subject in Lake Biandantang, a macrophytic lake in Hubei Province. Food webs of the macroinvertebrate community were compiled bimonthly from March, 2002 to March, 2003. Dietary information was obtained from gut analysis. Linkage strength was quantified by combining estimates of energy flow (secondary production) with data of gut analysis. The macroinvertebrate community of Lake Biandantang was based heavily on detritus. Quantitative food webs showed the total ingestion ranged from 6930 to 36,340 mg dry mass m(-2) bimonthly. The ingestion of macroinvertebrate community was higher in the months with optimum temperature than that in other periods with higher or lower temperature. Through comparison, many patterns in benthic food web of Lake Biandantang are consistent with other detritus-based webs, such as stream webs, but different greatly from those based on autochthonous primary production (e.g. pelagic systems). It suggests that the trophic basis of the web is essential in shaping food web structure.

Modelling of settlement induced building damage

This thesis focuses on the modelling of settlement induced damage to masonry buildings. In densely populated areas, the need for new space is nowadays producing a rapid increment of underground excavations. Due to the construction of new metro lines, tunnelling activity in urban areas is growing. One of the consequences is a greater attention to the risk of damage on existing structures. Thus, the assessment of potential damage of surface buildings has become an essential stage in the excavation projects in urban areas (Chapter 1). The current damage risk assessment procedure is based on strong simplifications, which not always lead to conservative results. Object of this thesis is the development of an improved damage classification system, which takes into account the parameters influencing the structural response to settlement, like the non-linear behaviour of masonry and the soil-structure interaction. The methodology used in this research is based on experimental and numerical modelling. The design and execution of an experimental benchmark test representative of the problem allows to identify the principal factors and mechanisms involved. The numerical simulations enable to generalize the results to a broader range of physical scenarios. The methodological choice is based on a critical review of the currently available procedures for the assessment of settlement-induced building damage (Chapter 2). A new experimental test on a 1/10th masonry façade with a rubber base interface is specifically designed to investigate the effect of soil-structure interaction on the tunnelling-induced damage (Chapter 3). The experimental results are used to validate a 2D semi-coupled finite element model for the simulation of the structural response (Chapter 4). The numerical approach, which includes a continuum cracking model for the masonry and a non-linear interface to simulate the soil-structure interaction, is then used to perform a sensitivity study on the effect of openings, material properties, initial damage, initial conditions, normal and shear behaviour of the base interface and applied settlement profile (Chapter 5). The results assess quantitatively the major role played by the normal stiffness of the soil-structure interaction and by the material parameters defining the quasi-brittle masonry behaviour. The limitation of the 2D modelling approach in simulating the progressive 3D displacement field induced by the excavation and the consequent torsional response of the building are overcome by the development of a 3D coupled model of building, foundation, soil and tunnel (Chapter 6). Following the same method applied to the 2D semi-coupled approach, the 3D model is validated through comparison with the monitoring data of a literature case study. The model is then used to carry out a series of parametric analyses on geometrical factors: the aspect ratio of horizontal building dimensions with respect to the tunnel axis direction, the presence of adjacent structures and the position and alignment of the building with respect to the excavation (Chapter 7). The results show the governing effect of the 3D building response, proving the relevance of 3D modelling. Finally, the results from the 2D and 3D parametric analyses are used to set the framework of an overall damage model which correlates the analysed structural features with the risk for the building of being damaged by a certain settlement (Chapter 8). This research therefore provides an increased experimental and numerical understanding of the building response to excavation-induced settlements, and sets the basis for an operational tool for the risk assessment of structural damage (Chapter 9).

Molecular basis of transferrin polymorphism in goldfish (Carassius auratus)

Transferrin (TF) polymorphism was investigated in a color variety of goldfish (Carassius auratus), and its molecular basis analyzed. Three TF variants (A(1), A(2) and B-1) were identified from an inbred strain of the goldfish, of which A(1) and B-1 displayed a large electrophoretic difference on both native and SDS-PAGE gels. The TF cDNAs corresponding to variants A(1) and B-1 were cloned and sequenced from A(1)A(1), A(1)B(1) and B1B1 individuals, and their deduced amino acid sequences were analyzed. Substantial amino acid variation occurred between variants A(1) and B-1, with significant differences in peptide length, theoretical molecular weight (Mw) and isoelectric point (pI). No potential glycosylation sites were observed in the two amino acid sequences, which excluded the possibility that carbohydrate difference might cause electrophoretic variation among the TF variants. Further analysis suggested that the distinct electrophoretic mobility of the two variants A(1) and B-1 by SDS-PAGE resulted from their Mw difference, while the difference by the native PAGE could be explained by their pI variation. Furthermore, genomic DNA fragments containing the transferrin alleles were amplified and subjected to RFLP analysis in A(1)A(1), A(1)B(1) and B1B1 individuals. The data revealed characteristic banding patterns for each TF genotype, and demonstrated that the TF alleles A(1) and B-1 could be used as a co-dominant marker system. The initial work relating to the goldfish TF variants will benefit the understanding of the evolutionary and functional significance of TF polymorphism in fish.

Fixed-rank matrix factorizations and Riemannian low-rank optimization

Motivated by the problem of learning a linear regression model whose parameter is a large fixed-rank non-symmetric matrix, we consider the optimization of a smooth cost function defined on the set of fixed-rank matrices. We adopt the geometric framework of optimization on Riemannian quotient manifolds. We study the underlying geometries of several well-known fixed-rank matrix factorizations and then exploit the Riemannian quotient geometry of the search space in the design of a class of gradient descent and trust-region algorithms. The proposed algorithms generalize our previous results on fixed-rank symmetric positive semidefinite matrices, apply to a broad range of applications, scale to high-dimensional problems, and confer a geometric basis to recent contributions on the learning of fixed-rank non-symmetric matrices. We make connections with existing algorithms in the context of low-rank matrix completion and discuss the usefulness of the proposed framework. Numerical experiments suggest that the proposed algorithms compete with state-of-the-art algorithms and that manifold optimization offers an effective and versatile framework for the design of machine learning algorithms that learn a fixed-rank matrix. © 2013 Springer-Verlag Berlin Heidelberg.

Study on Closed-set Speaker Identification Based on Biomimetic Pattern Recognition

In this paper, a new classifier of speaker identification has been proposed, which is based on Biomimetic pattern recognition (BPR). Distinguished from traditional speaker recognition methods, such as DWT, HMM, GMM, SVM and so on, the proposed classifier is constructed by some finite sub-space which is reasonable covering of the points in high dimensional space according to distributing characteristic of speech feature points. It has been used in the system of speaker identification. Experiment results show that better effect could be obtained especially with lesser samples. Furthermore, the proposed classifier employs a much simpler modeling structure as compared to the GMM. In addition, the basic idea "cognition" of Biomimetic pattern recognition (BPR) results in no requirement of retraining the old system for enrolling new speakers.

Biomimetic pattern recognition theory and its applications

Biomimetic pattern recogntion (BPR), which is based on "cognition" instead of "classification", is much closer to the function of human being. The basis of BPR is the Principle of homology-continuity (PHC), which means the difference between two samples of the same class must be gradually changed. The aim of BPR is to find an optimal covering in the feature space, which emphasizes the "similarity" among homologous group members, rather than "division" in traditional pattern recognition. Some applications of BPR are surveyed, in which the results of BPR are much better than the results of Support Vector Machine. A novel neuron model, Hyper sausage neuron (HSN), is shown as a kind of covering units in BPR. The mathematical description of HSN is given and the 2-dimensional discriminant boundary of HSN is shown. In two special cases, in which samples are distributed in a line segment and a circle, both the HSN networks and RBF networks are used for covering. The results show that HSN networks act better than RBF networks in generalization, especially for small sample set, which are consonant with the results of the applications of BPR. And a brief explanation of the HSN networks' advantages in covering general distributed samples is also given.

The application of DBF neural networks for object recognition

On the basis of DBF nets proposed by Wang Shoujue, the model and properties of DBF neural network were discussed in this paper. When applied in pattern recognition, the algorithm and implement on hardware were presented respectively. We did experiments on recognition of omnidirectionally oriented rigid objects on the same level, using direction basis function neural networks, which acts by the method of covering the high dimensional geometrical distribution of the sample set in the feature space. Many animal and vehicle models (even with rather similar shapes) were recognized omnidirectionally thousands of times. For total 8800 tests, the correct recognition rate is 98.75%, the error rate and the rejection rate are 0.5% and 1.25% respectively. (C) 2003 Elsevier Inc. All rights reserved.

Geometrical learning, descriptive geometry, and biomimetic pattern recognition

Studies on learning problems from geometry perspective have attracted an ever increasing attention in machine learning, leaded by achievements on information geometry. This paper proposes a different geometrical learning from the perspective of high-dimensional descriptive geometry. Geometrical properties of high-dimensional structures underlying a set of samples are learned via successive projections from the higher dimension to the lower dimension until two-dimensional Euclidean plane, under guidance of the established properties and theorems in high-dimensional descriptive geometry. Specifically, we introduce a hyper sausage like geometry shape for learning samples and provides a geometrical learning algorithm for specifying the hyper sausage shapes, which is then applied to biomimetic pattern recognition. Experimental results are presented to show that the proposed approach outperforms three types of support vector machines with either a three degree polynomial kernel or a radial basis function kernel, especially in the cases of high-dimensional samples of a finite size. (c) 2005 Elsevier B.V. All rights reserved.

Detecting a set of entanglement measures in an unknown tripartite quantum state by local operations and classical communication

We propose a more general method for detecting a set of entanglement measures, i.e., negativities, in an arbitrary tripartite quantum state by local operations and classical communication. To accomplish the detection task using this method, three observers do not need to perform partial transposition maps by the structural physical approximation; instead, they only need to collectively measure some functions via three local networks supplemented by a classical communication. With these functions, they are able to determine the set of negativities related to the tripartite quantum state.

Computer-aided design of PV/wind hybrid system

A complete set of match calculation methods for optimum sizing of PV/wind hybrid system is presented. In this method, the more accurate and practical mathematic models for characterizing PV module, wind generator and battery are adopted; combining with hourly measured meteorologic data and load data, the performance of a PV/wind hybrid system is determined on a hourly basis; by fixing the capacity of wind generators, the whole year's LPSP (loss of power supply probability) values of PV/wind hybrid systems with different capacity of PV array and battery bank are calculated, then the trade-off curve between battery bank and PV array capacity is drawn for the given LPSP value; the optimum configuration which can meet the energy demand with the minimum cost can be found by drawing a tangent to the trade-off curve with the slope representing the relationship between cost of PV module and that of the battery. According to this match calculation method, a set of match calculation programs for optimum sizing of PV/wind hybrid systems have been developed. Applying these match calculation programs to an assumed PV/wind hybrid system to be installed at Waglan island of Hong Kong, the optimum configuration and its hourly, daily, monthly and yearly performances are given. (C) 2003 Elsevier Science Ltd. All rights reserved.

An adaptive controller for a class of nonlinear system using direction basis function

For a class of nonlinear dynamical systems, the adaptive controllers are investigated using direction basis function (DBF) in this paper. Based on the criterion of Lyapunov' stability, DBF is designed which guarantees that the output of the controlled system asymptotically tracks the reference signals. Finally, the simulation shows the good tracking effectiveness of the adaptive controller.

Modeling and optimization of semiconductor manufacturing process with neural networks

A neural network-based process model is proposed to optimize the semiconductor manufacturing process. Being different from some works in several research groups which developed neural network-based models to predict process quality with a set of process variables of only single manufacturing step, we applied this model to wafer fabrication parameters control and wafer lot yield optimization. The original data are collected from a wafer fabrication line, including technological parameters and wafer test results. The wafer lot yield is taken as the optimization target. Learning from historical technological records and wafer test results, the model can predict the wafer yield. To eliminate the "bad" or noisy samples from the sample set, an experimental method was used to determine the number of hidden units so that both good learning ability and prediction capability can be obtained.

Cancer classification based on direction-basis-function neural networks

Automatic molecular classification of cancer based on DNA microarray has many advantages over conventional classification based on morphological appearance of the tumor. Using artificial neural networks is a general approach for automatic classification. In this paper, Direction-Basis-Function neuron and Priority-Ordered algorithm are applied to neural networks. And the leukemia gene expression dataset is used as an example to testify the classifier. The result of our method is compared to that of SVM. It shows that our method makes a better performance than SVM.

Analysis of leakage current in GaAs micro-solar cell arrays

The output characteristics of micro-solar cell arrays are analyzed on the basis of a modified model in which the shunt resistance between cell lines results in current leakage. The modification mainly consists of adding a shunt resistor network to the traditional model. The obtained results agree well with the reported experimental results. The calculation results demonstrate that leakage current in substrate affects seriously the performance of GaAs micro- solar cell arrays. The performance of arrays can be improved by reducing the number of cells per line. In addition, at a certain level of integration, an appropriate space occupancy rate of the single cell is recommended for ensuring high open circuit voltages, and it is more appropriate to set the rates at 80%-90% through the calculation.