High-rate vector quantization for detection

We investigate high-rate quantization for various detection and reconstruction loss criteria. A new distortion measure is introduced which accounts for global loss in best attainable binary hypothesis testing performance. The distortion criterion is related to the area under the receiver-operating-characteristic (ROC) curve. Specifically, motivated by Sanov's theorem, we define a performance curve as the trajectory of the pair of optimal asymptotic Type I and Type II error rates of the most powerful Neyman-Pearson test of the hypotheses. The distortion measure is then defined as the difference between the area-under-the-curve (AUC) of the optimal pre-encoded hypothesis test and the AUC of the optimal post-encoded hypothesis test. As compared to many previously introduced distortion measures for decision making, this distortion measure has the advantage of being independent of any detection thresholds or priors on the hypotheses, which are generally difficult to specify in the code design process. A high-resolution Zador-Gersho type of analysis is applied to characterize the point density and the inertial profile associated with the optimal high-rate vector quantizer. The analysis applies to a restricted class of high-rate quantizers that have bounded cells with vanishing volumes. The optimal point density is used to specify a Lloyd-type algorithm which allocates its finest resolution to regions where the gradient of the pre-encoded likelihood ratio has greatest magnitude.     

Information analysis of single photon emission computed tomography with count losses

An analysis is presented of the information transfer from emitter-space to detector-space in single photon emission computed tomography (SPECT) systems. The analysis takes into account the fact that count loss side information is generally not available at the detector. Side information corresponds to the number gamma-rays lost deleted due to lack of interaction with the detector data. It is shown that the information transfer depends on the structure of the likelihood function of the emitter locations associated with the detector data. This likelihood function is the average of a set of ideal-detection likelihood functions, each matched to a particular set of possible deleted gamma-ray paths. A lower bound is derived for the information gain due to incorporating the count loss side information at the detector. This is shown to be significant when the mean emission rate is small or when the gamma-ray deletion probability is strongly dependent on emitter location. Numerical evaluations of the mutual information, with and without side information, associated with information-optimal apertures and uniform parallel-hole collimators are presented.     

A regularized graph layout framework for dynamic network visualization

Many real-world networks, including social and information networks, are dynamic structures that evolve over time. Such dynamic networks are typically visualized using a sequence of static graph layouts. In addition to providing a visual representation of the network structure at each time step, the sequence should preserve the mental map between layouts of consecutive time steps to allow a human to interpret the temporal evolution of the network. In this paper, we propose a framework for dynamic network visualization in the on-line setting where only present and past graph snapshots are available to create the present layout. The proposed framework creates regularized graph layouts by augmenting the cost function of a static graph layout algorithm with a grouping penalty, which discourages nodes from deviating too far from other nodes belonging to the same group, and a temporal penalty, which discourages large node movements between consecutive time steps. The penalties increase the stability of the layout sequence, thus preserving the mental map. We introduce two dynamic layout algorithms within the proposed framework, namely dynamic multidimensional scaling and dynamic graph Laplacian layout. We apply these algorithms on several data sets to illustrate the importance of both grouping and temporal regularization for producing interpretable visualizations of dynamic networks.     

减少离散大系统两级镇定控制的保守性

The aim of this paper is to propose a new algorithm for multilevel stabilization of large scale systems. In two-level stabilization method, a set of local stabilizers for the individual subsystems in a completely decentralized environment is designed. The solution of the control problem involves designing of a global controller on a higher hierarchical level that provides corrective signals to account for interconnections effect. The principle feature of this paper is to reduce conservativeness in global controller design. Here, the key point is to reduce the effect of interactions instead of neutralizing them. In fact, unlike prior methods, our idea does not ignore the possible beneficial aspects of the interactions and does not try to neutralize them.     

Solving nonlinear optimal control problems using a hybrid IPSO-SQP algorithm

A hybrid algorithm by integrating an improved particle swarm optimization (IPSO) with successive quadratic programming (SQP), namely IPSO-SQP, is proposed for solving nonlinear optimal control problems. The particle swarm optimization (PSO) is showed to converge rapidly to a near optimum solution, but the search process will become very slow around global optimum. On the contrary, the ability of SQP is weak to escape local optimum but can achieve faster convergent speed around global optimum and the convergent accuracy can be higher. Hence, in the proposed method, at the beginning stage of search process, a PSO algorithm is employed to find a near optimum solution. In this case, an improved PSO (IPSO) algorithm is used to enhance global search ability and convergence speed of algorithm. When the change in fitness value is smaller than a predefined value, the searching process is switched to SQP to accelerate the search process and find an accurate solution. In this way, this hybrid algorithm may find an optimum solution more accurately. To validate the performance of the proposed IPSO-SQP approach, it is evaluated on two optimal control problems. Results show that the performance of the proposed algorithm is satisfactory.     

A general insight into the effect of neuron structure on classification

This paper gives a general insight into how the neuron structure in a multilayer perceptron (MLP) can affect the ability of neurons to deal with classification. Most of the common neuron structures are based on monotonic activation functions and linear input mappings. In comparison, the proposed neuron structure utilizes a nonmonotonic activation function and/or a nonlinear input mapping to increase the power of a neuron. An MLP of these high power neurons usually requires a less number of hidden nodes than conventional MLP for solving classification problems. The fewer number of neurons is equivalent to the smaller number of network weights that must be optimally determined by a learning algorithm. The performance of learning algorithm is usually improved by reducing the number of weights, i.e., the dimension of the search space. This usually helps the learning algorithm to escape local optimums, and also, the convergence speed of the algorithm is increased regardless of which algorithm is used for learning. Several 2-dimensional examples are provided manually to visualize how the number of neurons can be reduced by choosing an appropriate neuron structure. Moreover, to show the efficiency of the proposed scheme in solving real-world classification problems, the Iris data classification problem is solved using an MLP whose neurons are equipped by nonmonotonic activation functions, and the result is compared with two well-known monotonic activation functions.     

A Fast Spectral Method for Active 3D Shape Reconstruction

Variational energy minimization techniques for surface reconstruction are implemented by evolving an active surface according to the solutions of a sequence of elliptic partial differential equations (PDE's). For these techniques, most current approaches to solving the elliptic PDE are iterative involving the implementation of costly finite element methods (FEM) or finite difference methods (FDM). The heavy computational cost of these methods makes practical application to 3D surface reconstruction burdensome. In this paper, we develop a fast spectral method which is applied to 3D active surface reconstruction of star-shaped surfaces parameterized in polar coordinates. For this parameterization the Euler-Lagrange equation is a Helmholtz-type PDE governing a diffusion on the unit sphere. After linearization, we implement a spectral non-iterative solution of the Helmholtz equation by representing the active surface as a double Fourier series over angles in spherical coordinates. We show how this approach can be extended to include region-based penalization. A number of 3D examples and simulation results are presented to illustrate the performance of our fast spectral active surface algorithms.     

Synthesis and properties of cross‐linked polymers from epoxidized rubber seed oil and triethylenetetramine

A series of epoxidized oils were prepared from rubber seed, soybean, jatropha, palm, and coconut oils. The epoxy content varied from 0.03 to 7.4 wt %, in accordance with the degree of unsaturation of the oils (lowest for coconut, highest for rubber seed oil). Bulk polymerization/curing of the epoxidized oils with triethylenetetramine (in the absence of a catalyst) was carried out in a batch setup (1 : 1 molar ratio of epoxide to primary amine groups, 100°C, 100 rpm, 30 min) followed by casting of the mixture in a steel mold (180°C, 200 bar, 21 h) and this resulted in cross‐linked resins. The effect of relevant pressing conditions such as time, temperature, pressure, and molar ratio of the epoxide and primary amine groups was investigated and modeled using multivariable nonlinear regression. Good agreement between experimental data and model were obtained. The rubber seed oil‐derived polymer has a T_g of 11.1°C, a tensile strength of 1.72 MPa, and strain at break of 182%. These values are slightly higher than for commercial epoxidized soybean oil (T_g of 6.9°C, tensile strength of 1.11 MPa, and strain at break of 145.7%). However, the comparison highlights the potential for these novel resins to be used at industrial/commercial level. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 42591.     

Structural and piezoelectric properties of textured NLKNS-CZ thick films and their application in planar piezoactuator

0.97(Na_0.5-xLi_xK_0.5)(Nb_0.89Sb_0.11)O₃-0.03CaZrO₃ [(N_0.5-xL_xK)(NS)-CZ] piezoceramic (x = 0.325) has a pseudocubic-tetragonal-orthorhombic (PC-T-O) multi-structure. The PC structure formed in this piezoceramic was identified as the R3m rhombohedral structure. This piezoceramic showed the large piezoelectric charge constant (d₃₃) of 515 pC/N due to the PC-T-O multi-structure. The NaNbO₃ (NN) templates were used to texture the (N_0.5-xL_xK)(NS)-CZ thick films along the (001) direction, and the textured thick film (x = 0.0375) had a large Lotgering factor of 95.6%. The PC-T-O multi-structure was observed in this thick film (x = 0.0375), but the thick film (x = 0.0325) showed a PC-O structure owing to the diffusion of the NN templates into the thick film. The textured thick film (x = 0.0375) exhibited an increased d₃₃ of 625 pC/N because of the PC-T-O multi-structure and the lineup of grains along the [001] direction. A textured thick film (x = 0.0375) was used to fabricate a planar-type actuator to confirm its applicability to electrical devices. This actuator exhibits large acceleration (580.3 G) and displacement (150 μm) at a low electric field of 0.2 kV/mm with a short response time of 3.0 ms. Therefore, the (N_0.5-xL_xK)(NS)-CZ thick films are excellent lead-free piezoceramics.