A close relationship between the class of totally positive matrices and anti-Monge matrices is used for suggesting a new direction for investigating totally positive matrices. Some questions are posed and a partial answer in the case of Vandermonde-like matrices is given., Miroslav Fiedler., and Obsahuje seznam literatury
Combinatorial optimization is a discipline of decision making in the case of diserete alternatives. The Genetic Neighborhood Search (GNS) is a hybrid method for these combinatorial optimization problems. The main feature of the approach is iterative use of local search on extended neighborhoods, where the better solution will be the center of a new extended neighborhood. When the center of the neighborhood would be t.he better solution the algorithm will stop. We propose using a genetic algorithm to exi)lore the extended neighborhoods. This GA is characterized by the method of evaluating the fitness of individuals and useing two new operators. Computational experience with the Symmetric TSP shows that this approach is robust with respect to the starting point and that high quality solutions are obtained in a reasonable time.
The presented technological procedure makes it possible to assemble large magnetic blocks from permanent magnets with a high value of maximum energy product in such a way that the individual magnets or magnetic plates are moved toward each other at a controlled speed in the direction perpendicular to the future common contact surface of these magnets, i.e. parallel to the induction lines crossing this contact surface. Unlike in the previously used way of assembling the blocks, it is thus possible to eliminate the influence of partial demagnetization as the blocks are being assembled and consequently to reach higher values of magnetic induction in the air gap of the magnetic circuit. When applying the new method of assembling the blocks for instance in circuits of magnetic filters for the purification of ceramic suspensions, a prerequisite for the further improvement of the technological parameters of filtration is thus created., Václav Žežulka and Pavel Straka., and Obsahuje bibliografické odkazy
An unnamed microcercous cercaria (Digenea: Monorchiidae), a parasite of Amiantis purpurata (Lamarck, 1818) (Bivalvia: Veneridae) and its corresponding metacercaria from the province of Buenos Aires and the Patagonian coast of the Southwest Atlantic Ocean, are described. The cercaria described in this paper differs from the three other monorchiid microcercous cercariae, i.e., Lasiotocus minutus (Manter, 1931), Lasiotocus elongatus (Manter, 1931), and Cercaria caribbea XXXVI Cable, 1956, mainly because of the extension of the excretory vesicle and the location of the ventral sucker. Cercariae artificially extracted from sporocysts encyst in a dish and form metacercariae enveloped by a gelatinous sac with two prolongations, which are used to adhere to the substratum. The monorchiid described in this paper has a life cycle similar to those of L. minutus and L. elongatus, although the adult stage of the present species is still unknown. Their larvae are similar in morphology and have venerid clams as their first hosts. The presence of a monorchiid larva is reported for the first time in the Southern Hemisphere. Its monthly prevalence rates, ranging from 0 to 25% (mean: 8.3%), are given from the Patagonian coast. The infection seems to cause castration as it was observed that during March through to May, when most gametes were produced in uninfected individuals, 81% of the infected individuals did not produce gametes.
The multilayer perceptron model has been suggested as an alternative to conventional approaches, and can accurately forecast time series. Additionally, several novel artificial neural network models have been proposed as alternatives to the multilayer perceptron model, which have used (for example) the generalized- mean, geometric mean, and multiplicative neuron models. Although all of these artificial neural network models can produce successful forecasts, their aggregation functions mean that they are negatively affected by outliers. In this study, we propose a new multilayer, feed forward neural network model, which is a robust model that uses the trimmed mean neuron model. Its aggregation function does not depend on outliers. We trained this multilayer, feed forward neural network using modied particle swarm optimization. We applied the proposed method to three well-known time series, and our results suggest that it produces superior forecasts when compared with similar methods.
Neural networks have shown good results for detecting a certain pattern in a given image. In this paper, faster neural networks for pattern detection are presented. Such processors are designed based on cross correlation in the frequency domain between the input matrix and the input weights of neural networks. This approach is developed to reduce the computation steps required by these faster neural networks for the searching process. The principle of divide and conquer strategy is applied through matrix decomposition. Each matrix is divided into submatrices small in size, and then each one is tested separately by using a single faster neural processor. Furthermore, faster pattern detection is obtained by using parallel processing techniques to test the resulting submatrices at the same time, employing the same number of faster neural networks. In contrast to faster neural networks, the speed-up ratio is increased with the size of the input matrix when using faster neural networks and matrix decomposition. Moreover, the problem of local submatrix normalization in the frequency domain is solved. The effect of matrix normalization on the speed-up ratio of pattern detection is discussed. Simulation results show that local submatrix normalization through weight normalization is faster than submatrix normalization in the spatial domain. The overall speed-up ratio of the detection process is increased as the normalization of weights is done offline.
A new model for propagation of long waves including the coastal area is introduced. This model considers only the motion of the surface of the sea under the condition of preservation of mass and the sea floor is inserted into the model as an obstacle to the motion. Thus we obtain a constrained hyperbolic free-boundary problem which is then solved numerically by a minimizing method called {\em the discrete Morse semi-flow}. The results of the computation in 1D show the adequacy of the proposed model.
An original Nyquist-based frequency domain robust decentralized controller (DC) design technique for robust stability and guaranteed nominal performance is proposed, applicable for continuous-time uncertain systems described by a set of transfer function matrices. To provide nominal performance, interactions are included in individual design using one selected characteristic locus of the interaction matrix, used to reshape frequency responses of decoupled subsystems; such modified subsystems are termed "equivalent subsystems". Local controllers of equivalent subsystems independently tuned for stability and specified feasible performance constitute the decentralized controller guaranteeing specified performance of the full system. To guarantee robust stability, the M−Δ stability conditions are derived. Unlike standard robust approaches, the proposed technique considers full nominal model, thus reducing conservativeness of resulting robust stability conditions. The developed frequency domain design procedure is graphical, interactive and insightful. A case study providing a step-by-step robust DC design for the Quadruple Tank Process [K.H. Johansson: Interaction bounds in multivariable control systems. Automatica 38 (2002), 1045-1051] is included.
Community detection algorithms help us improve the management of complex networks and provide a clean sight of them. We can encounter complex networks in various fields such as social media, bioinformatics, recommendation systems, and search engines. As the definition of the community changes based on the problem considered, there is no algorithm that works universally for all kinds of data and network structures. Communities can be disjointed such that each member is in at most one community or overlapping such that every member is in at least one community. In this study, we examine the problem of finding overlapping communities in complex networks and propose a new algorithm based on the similarity of neighbors. This algorithm runs in O(mlgm) running time in the complex network containing m number of relationships. To compare our algorithm with existing ones, we select the most successful four algorithms from the Community Detection library (CDlib) by eliminating the algorithms that require prior knowledge, are unstable, and are time-consuming. We evaluate the successes of the proposed algorithm and the selected algorithms using various known metrics such as modularity, F-score, and Normalized Mutual Information. In addition, we adapt the coverage metric defined for disjoint communities to overlapping communities and also make comparisons with this metric. We also test all of the algorithms on small graphs of real communities. The experimental results show that the proposed algorithm is successful in finding overlapping communities.