Journal articles on the topic 'Network function configuration'
To see the other types of publications on this topic, follow the link: Network function configuration.
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Network function configuration.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Schmidt, A., C. Kruse, F. Rottensteiner, U. Soergel, and C. Heipke. "NETWORK DETECTION IN RASTER DATA USING MARKED POINT PROCESSES." ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLI-B3 (June 10, 2016): 701–8. http://dx.doi.org/10.5194/isprs-archives-xli-b3-701-2016.
Full textAbstract:
We propose a new approach for the automatic detection of network structures in raster data. The model for the network structure is represented by a graph whose nodes and edges correspond to junction-points and to connecting line segments, respectively; nodes and edges are further described by certain parameters. We embed this model in the probabilistic framework of marked point processes and determine the most probable configuration of objects by stochastic sampling. That is, different graph configurations are constructed randomly by modifying the graph entity parameters, by adding and removing nodes and edges to/ from the current graph configuration. Each configuration is then evaluated based on the probabilities of the changes and an energy function describing the conformity with a predefined model. By using the Reversible Jump Markov Chain Monte Carlo sampler, a global optimum of the energy function is determined. We apply our method to the detection of river and tidal channel networks in digital terrain models. In comparison to our previous work, we introduce constraints concerning the flow direction of water into the energy function. Our goal is to analyse the influence of different parameter settings on the results of network detection in both, synthetic and real data. Our results show the general potential of our method for the detection of river networks in different types of terrain.
2
Schmidt, A., C. Kruse, F. Rottensteiner, U. Soergel, and C. Heipke. "NETWORK DETECTION IN RASTER DATA USING MARKED POINT PROCESSES." ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLI-B3 (June 10, 2016): 701–8. http://dx.doi.org/10.5194/isprsarchives-xli-b3-701-2016.
Full textAbstract:
We propose a new approach for the automatic detection of network structures in raster data. The model for the network structure is represented by a graph whose nodes and edges correspond to junction-points and to connecting line segments, respectively; nodes and edges are further described by certain parameters. We embed this model in the probabilistic framework of marked point processes and determine the most probable configuration of objects by stochastic sampling. That is, different graph configurations are constructed randomly by modifying the graph entity parameters, by adding and removing nodes and edges to/ from the current graph configuration. Each configuration is then evaluated based on the probabilities of the changes and an energy function describing the conformity with a predefined model. By using the Reversible Jump Markov Chain Monte Carlo sampler, a global optimum of the energy function is determined. We apply our method to the detection of river and tidal channel networks in digital terrain models. In comparison to our previous work, we introduce constraints concerning the flow direction of water into the energy function. Our goal is to analyse the influence of different parameter settings on the results of network detection in both, synthetic and real data. Our results show the general potential of our method for the detection of river networks in different types of terrain.
3
Billings, Steve A., and Guang L. Zheng. "Radial basis function network configuration using genetic algorithms." Neural Networks 8, no. 6 (January 1995): 877–90. http://dx.doi.org/10.1016/0893-6080(95)00029-y.
Full text
4
Jaimes-Reátegui, R., J. M. Castillo-Cruz, J. H. Garcıa-Lopez, G. Huerta-Cuellar, L. A. Gallegos-Infante, and A. N. Pisarchik. "Bistability in network motifs of Duffing oscillators." Cybernetics and Physics, Volume 9, 2020, Number 1 (June 30, 2020): 31–40. http://dx.doi.org/10.35470/2226-4116-2020-9-1-31-40.
Full textAbstract:
We study the emergence of synchronization in the network motif of three bistable Duffing oscillators coupled in all possible configurations. The equation of motion is derived for every configuration. For each motif, we vary initial conditions of every oscillator and calculate the bifurcation diagram as a function of the coupling strength. We find transitions of the whole system to a monostable regime with either a fixed point or a limit cycle depending on the motif’s configuration, as the coupling strength is increased. The most complex dynamics is observed the nidirectional chain, where a transition to quasiperiodicity occurs.
5
Mateichyk, Vasyl, Miroslaw Śmieszek, and Nataliia Kostian. "Evaluation of transport system configuration by efficiency indicators." Transport technologies 2022, no. 2 (December 10, 2022): 52–62. http://dx.doi.org/10.23939/tt2022.02.052.
Full textAbstract:
The study is devoted to the process of evaluating the efficiency of the transport system in terms of urban mobility. The approach is based on the use of a system of performance indicators using neurocomputer technologies. Generalized models for obtaining a vector of performance indicators and an integral performance indicator in the form of computer neural networks are proposed. It is shown that to record the fact that the indicator values fall to the threshold and below, it is enough to use a neural network built on perceptron neurons. The multi-layered model for determining the integral indicator allows assessing the importance of individual indicators in the system of monitoring the efficiency of a given configuration of the transport system. An experimental study of twenty-five states of the transport system of various configurations in the cities of Poland and Ukraine was carried out. The key indicators of the system's efficiency are determined, namely, the energy efficiency indicator of the vehicle as a system element, the environmental indicator and the traffic safety indicator. Based on the results of the experimental study, a neural network structure is proposed for evaluating the energy efficiency of given configurations of the transport system. For the purpose of training and testing the obtained network, the procedure of adjusting the threshold value of the activation function and normalizing the values of the input parameters array of the transport system was used. The constructed network was implemented using Visual Studio 2019 using the C++ language. The network was adjusted to determine the energy efficiency estimate with a given accuracy by replacing the perceptron neuron with a regular one with a sigmoidal activation function. The random nature of the choice of the configuration and the initial values of the weighting factors made it possible to obtain a model with an accuracy of implementation on the control sample in the range from 90 to 98.7% at a learning rate of 0.1.
6
SHARPE, ROBERT N., and MO-YUEN CHOW. "RELATIONSHIP BETWEEN A FUZZY LOGIC AND A STEEPEST DESCENT APPROACH TO OPTIMIZE A FEEDFORWARD ARTIFICIAL NEURAL NETWORK CONFIGURATION." International Journal of Neural Systems 05, no. 04 (December 1994): 299–312. http://dx.doi.org/10.1142/s012906579400030x.
Full textAbstract:
The neural network designer must take into consideration many factors when selecting an appropriate network configuration. The performance of a given network configuration is influenced by many different factors such as: accuracy, training time, sensitivity, and the number of neurons used in the implementation. Using a cost function based on the four criteria mentioned previously, the various network paradigms can be evaluated relative to one another. If the mathematical models of the evaluation criteria as functions of the network configuration are known, then traditional techniques (such as the steepest descent method) could be used to determine the optimal network configuration. The difficulty in selecting an appropriate network configuration is due to the difficulty involved in determining the mathematical models of the evaluation criteria. This difficulty can be avoided by using fuzzy logic techniques to perform the network optimization as opposed to the traditional techniques. Fuzzy logic avoids the need of a detailed mathematical description of the relationship between the network performance and the network configuration, by using heuristic reasoning and linguistic variables. A comparison will be made between the fuzzy logic approach and the steepest descent method for the optimization of the cost function. The fuzzy optimization procedure could be applied to other systems where there is a priori information about their characteristics.
7
Rădulescu, Anca. "Neural Network Spectral Robustness under Perturbations of the Underlying Graph." Neural Computation 28, no. 1 (January 2016): 1–44. http://dx.doi.org/10.1162/neco_a_00798.
Full textAbstract:
Recent studies have been using graph-theoretical approaches to model complex networks (such as social, infrastructural, or biological networks) and how their hardwired circuitry relates to their dynamic evolution in time. Understanding how configuration reflects on the coupled behavior in a system of dynamic nodes can be of great importance, for example, in the context of how the brain connectome is affecting brain function. However, the effect of connectivity patterns on network dynamics is far from being fully understood. We study the connections between edge configuration and dynamics in a simple oriented network composed of two interconnected cliques (representative of brain feedback regulatory circuitry). In this article our main goal is to study the spectra of the graph adjacency and Laplacian matrices, with a focus on three aspects in particular: (1) the sensitivity and robustness of the spectrum in response to varying the intra- and intermodular edge density, (2) the effects on the spectrum of perturbing the edge configuration while keeping the densities fixed, and (3) the effects of increasing the network size. We study some tractable aspects analytically, then simulate more general results numerically, thus aiming to motivate and explain our further work on the effect of these patterns on the network temporal dynamics and phase transitions. We discuss the implications of such results to modeling brain connectomics. We suggest potential applications to understanding synaptic restructuring in learning networks and the effects of network configuration on function of regulatory neural circuits.
8
LIU, YABO, JIANHUA YANG, and ZHAOHUI WU. "UBIQUITOUS AND COOPERATIVE NETWORK ROBOT SYSTEM WITHIN A SERVICE FRAMEWORK." International Journal of Humanoid Robotics 08, no. 01 (March 2011): 147–67. http://dx.doi.org/10.1142/s021984361100237x.
Full textAbstract:
Network robot system (NRS) is a new concept that integrates physical autonomous robots, environmental sensors, and human–robot interactions through network-based cooperation. The aim of this paper is to provide a ubiquitous and cooperative service framework for NRS. We first present foundational concepts of semantic map and service definition for the framework. Then, in order to generate feasible service configurations to fulfill tasks, we propose service configuration and reconfiguration algorithms, which dynamically search the appropriate service configurations for different tasks. Additionally, we put forward a service reasoning and enabling process to tackle the service unavailable problems. A cost evaluation function for service configuration is also proposed to facilitate the selection of suitable configurations. We tested and evaluated the framework in both simulation system and physical environment. Specifically, by separately varying the parameter settings, system performance was measured in three aspects: the success rate of tasks, the average waiting time per task, and the average cost per task. The experiment results indicate that the versatile service framework provides self-adaptive capability and utilizes available resources efficiently under a range of different scenarios.
9
Bonkhoff, Anna K., Flor A. Espinoza, Harshvardhan Gazula, Victor M. Vergara, Lukas Hensel, Jochen Michely, Theresa Paul, et al. "Acute ischaemic stroke alters the brain’s preference for distinct dynamic connectivity states." Brain 143, no. 5 (May 1, 2020): 1525–40. http://dx.doi.org/10.1093/brain/awaa101.
Full textAbstract:
Abstract Acute ischaemic stroke disturbs healthy brain organization, prompting subsequent plasticity and reorganization to compensate for the loss of specialized neural tissue and function. Static resting state functional MRI studies have already furthered our understanding of cerebral reorganization by estimating stroke-induced changes in network connectivity aggregated over the duration of several minutes. In this study, we used dynamic resting state functional MRI analyses to increase temporal resolution to seconds and explore transient configurations of motor network connectivity in acute stroke. To this end, we collected resting state functional MRI data of 31 patients with acute ischaemic stroke and 17 age-matched healthy control subjects. Stroke patients presented with moderate to severe hand motor deficits. By estimating dynamic functional connectivity within a sliding window framework, we identified three distinct connectivity configurations of motor-related networks. Motor networks were organized into three regional domains, i.e. a cortical, subcortical and cerebellar domain. The dynamic connectivity patterns of stroke patients diverged from those of healthy controls depending on the severity of the initial motor impairment. Moderately affected patients (n = 18) spent significantly more time in a weakly connected configuration that was characterized by low levels of connectivity, both locally as well as between distant regions. In contrast, severely affected patients (n = 13) showed a significant preference for transitions into a spatially segregated connectivity configuration. This configuration featured particularly high levels of local connectivity within the three regional domains as well as anti-correlated connectivity between distant networks across domains. A third connectivity configuration represented an intermediate connectivity pattern compared to the preceding two, and predominantly encompassed decreased interhemispheric connectivity between cortical motor networks independent of individual deficit severity. Alterations within this third configuration thus closely resembled previously reported ones originating from static resting state functional MRI studies post-stroke. In summary, acute ischaemic stroke not only prompted changes in connectivity between distinct networks, but it also caused characteristic changes in temporal properties of large-scale network interactions depending on the severity of the individual deficit. These findings offer new vistas on the dynamic neural mechanisms underlying acute neurological symptoms, cortical reorganization and treatment effects in stroke patients.
10
Engelmann, Anna, and Admela Jukan. "A Combinatorial Reliability Analysis of Generic Service Function Chains in Data Center Networks." ACM Transactions on Modeling and Performance Evaluation of Computing Systems 6, no. 3 (September 30, 2021): 1–24. http://dx.doi.org/10.1145/3477046.
Full textAbstract:
In data center networks, the reliability of Service Function Chain (SFC)—an end-to-end service presented by a chain of virtual network functions (VNFs)—is a complex and specific function of placement, configuration, and application requirements, both in hardware and software. Existing approaches to reliability analysis do not jointly consider multiple features of system components, including, (i) heterogeneity, (ii) disjointness, (iii) sharing, (iv) redundancy, and (v) failure interdependency. To this end, we develop a novel analysis of service reliability of the so-called generic SFC, consisting of n = k + r sub-SFCs, whereby k ≥ 1 and r ≥ 0 are the numbers of arbitrary placed primary and backup (redundant) sub-SFCs, respectively. Our analysis is based on combinatorics and a reduced binomial theorem—resulting in a simple approach, which, however, can be utilized to analyze rather complex SFC configurations. The analysis is practically applicable to various VNF placement strategies in arbitrary data center configurations, and topologies and can be effectively used for evaluation and optimization of reliable SFC placements.
11
Ma, Zheng, Guang Quan Wang, and Jun Jie Xia. "The Brief Analysis of Solutions to Automated Network Equipment Security Configuration Baseline Verification." Applied Mechanics and Materials 713-715 (January 2015): 2360–65. http://dx.doi.org/10.4028/www.scientific.net/amm.713-715.2360.
Full textAbstract:
This paper introduces definition, control object and classification of network equipment security configuration baseline, analyses the difficulties of security work of the security configuration baseline, discusses function demand, the system architecture and the deployment module of the security configuration baseline verification , forms basic shape of the solutions to the security configuration baseline verification and gives suggestions to improve the security configuration baseline indemnification.
12
Prakash, K., F. R. Islam, K. A. Mamun, and H. R. Pota. "Configurations of Aromatic Networks for Power Distribution System." Sustainability 12, no. 10 (May 25, 2020): 4317. http://dx.doi.org/10.3390/su12104317.
Full textAbstract:
A distribution network is one of the main parts of a power system that distributes power to customers. While there are various types of power distribution networks, a recently introduced novel structure of an aromatic network could begin a new era in the distribution levels of power systems and designs of microgrids or smart grids. In order to minimize blackout periods during natural disasters and provide sustainable energy, improve energy efficiency and maintain stability of a distribution network, it is essential to configure/reconfigure the network topology based on its geographical location and power demand, and also important to realize its self-healing function. In this paper, a strategy for reconfiguring aromatic networks based on structures of natural aromatic molecules is explained. Various network structures are designed, and simulations have been conducted to justify the performance of each configuration. It is found that an aromatic network does not need to be fixed in a specific configuration (i.e., a DDT structure), which provides flexibility in designing networks and demonstrates that the successful use of such structures will be a perfect solution for both distribution networks and microgrid systems in providing sustainable energy to the end users.
13
Tam, Prohim, Sa Math, and Seokhoon Kim. "Priority-Aware Resource Management for Adaptive Service Function Chaining in Real-Time Intelligent IoT Services." Electronics 11, no. 19 (September 20, 2022): 2976. http://dx.doi.org/10.3390/electronics11192976.
Full textAbstract:
The growth of the Internet of Things (IoT) in various mission-critical applications generates service heterogeneity with different priority labels. A set of virtual network function (VNF) orders represents service function chaining (SFC) for a particular service to robustly execute in a network function virtualization (NFV)-enabled environment. In IoT networks, the configuration of adaptive SFC has emerged to ensure optimality and elasticity of resource expenditure. In this paper, priority-aware resource management for adaptive SFC is provided by modeling the configuration of real-time IoT service requests. The problem models of the primary features that impact the optimization of configuration times and resource utilization are studied. The proposed approaches query the promising embedded deep reinforcement learning engine in the management layer (e.g., orchestrator) to observe the state features of VNFs, apply the action on instantiating and modifying new/created VNFs, and evaluate the average transmission delays for end-to-end IoT services. In the embedded SFC procedures, the agent formulates the function approximator for scoring the existing chain performance metrics. The testbed simulation was conducted in SDN/NFV topologies and captured the average of rewards, delays, delivery ratio, and throughput as −48.6666, 10.9766 ms, 99.9221%, and 615.8441 Mbps, which outperformed other reference approaches, following parameter configuration in this environment.
14
Elagin, Vasily S., Alexander V. Bogachev, and Ilya A. Belozertsev. "Modeling the estimation of end-to-end packet latency for a chain of NFV nodes in 5G networks." T-Comm 16, no. 3 (2022): 23–30. http://dx.doi.org/10.36724/2072-8735-2022-16-3-23-30.
Full textAbstract:
It is expected that future communication networks will provide configurable delay-sensitive types of services (for example, streaming video, machine interaction). To support a variety of applications and use cases of servers providing various functions, you can use network function virtualization (NFV), which will be able to provide flexible implementation and placement of configuration of the necessary network functions. This article analyzes the end-to-end packet latency (E2E) for multiple traffic flows passing through the chain of embedded virtual network functions (VNF) in fifth-generation communication networks (5G). The Dominant of Generalized Resource Processing (DR-GPS) is used to distribute computing resources and transfer data between threads in each node of Network Function Virtualization (NFV) to achieve equitable distribution and utilization of available resources. The tandem queuing model is designed for incoming packets combined in several streams passing through the NFV node and its outgoing transmission channel. To analyze manageability, we separate packet processing (and transmission) of various streams in the simulation and determine the average packet processing and transmission rates of each stream as approximate service speeds.
15
Her, Oon Yi, Mohd Saiful Azimi Mahmud, Mohamad Shukri Zainal Abidin, Razman Ayop, and Salinda Buyamin. "Artificial neural network based short term electrical load forecasting." International Journal of Power Electronics and Drive Systems (IJPEDS) 13, no. 1 (March 1, 2022): 586. http://dx.doi.org/10.11591/ijpeds.v13.i1.pp586-593.
Full textAbstract:
In power generation, a 24-hour load profile can vary significantly throughout the day. Therefore, power generation must be adjusted to reduce money loss due to excess generation. This paper presents a short-term load forecasting (STLF) system design using artificial neural network (ANN). As ANN come in many different configurations, this paper analyzes the best ANN configuration via trial-and-error method. To train the ANN, historical load data from 2016 to 2018 of power south energy cooperative (AEC) is used. A simple feedforward ANN type with one hidden layer is implemented, where 48 neurons are used at the input layer. For hidden layer, an arbitrary 50 neurons are chosen and 24 neurons at output layer are used to generate a day ahead 24-hour load profile. To measure the best activation function for SLTF application, four non-linear activation functions will be tested and the best activation function is used to create two and three hidden layer ANN architecture. Finally, the performance of the two new networks will be compared against one hidden layer model. From the obtained result, the best performing model is found as two hidden layers ANN with Tanh as its hidden layer activation function with 8.9% of testing mean absolute percentage error (MAPE).
16
Zhang, Yin Juan. "Research and Application of OPC Network Communication in Configuration Software." Advanced Materials Research 774-776 (September 2013): 1774–77. http://dx.doi.org/10.4028/www.scientific.net/amr.774-776.1774.
Full textAbstract:
The network communication between MATLAB and configuration software is established using OPC (OLE for Process Control) technology, which is easily to give full play to generate interactive interface in configuration software and simulation of control object model in MATLAB software. The calculation function of weakness in configuration software is repaired and the control object model and controller is to be isolated. The effective simulation platform for theory research and design of control system is established.
17
Zhang, Yin Juan. "Research and Application of DDE Network Communication in Configuration Software." Applied Mechanics and Materials 401-403 (September 2013): 1935–38. http://dx.doi.org/10.4028/www.scientific.net/amm.401-403.1935.
Full textAbstract:
The network communication between MATLAB and configuration software is established using DDE(Dynamic Data Exchange) technology, which is easily to give full play to generate interactive interface in configuration software and simulation of control object model in MATLAB software. The calculation function of weakness in configuration software is repaired and the control object model and controller is to be isolated. The effective simulation platform for theory research and design of control system is established.
18
Luo, Guoliang, Bingqin He, Yanbo Xiong, Luqi Wang, Hui Wang, Zhiliang Zhu, and Xiangren Shi. "An Optimized Convolutional Neural Network for the 3D Point-Cloud Compression." Sensors 23, no. 4 (February 16, 2023): 2250. http://dx.doi.org/10.3390/s23042250.
Full textAbstract:
Due to the tremendous volume taken by the 3D point-cloud models, knowing how to achieve the balance between a high compression ratio, a low distortion rate, and computing cost in point-cloud compression is a significant issue in the field of virtual reality (VR). Convolutional neural networks have been used in numerous point-cloud compression research approaches during the past few years in an effort to progress the research state. In this work, we have evaluated the effects of different network parameters, including neural network depth, stride, and activation function on point-cloud compression, resulting in an optimized convolutional neural network for compression. We first have analyzed earlier research on point-cloud compression based on convolutional neural networks before designing our own convolutional neural network. Then, we have modified our model parameters using the experimental data to further enhance the effect of point-cloud compression. Based on the experimental results, we have found that the neural network with the 4 layers and 2 strides parameter configuration using the Sigmoid activation function outperforms the default configuration by 208% in terms of the compression-distortion rate. The experimental results show that our findings are effective and universal and make a great contribution to the research of point-cloud compression using convolutional neural networks.
19
Cherkasov, Dmytro. "Defining the Rules and Basic Set of Funtional Elements for Effective Modeling of Communication Networks." NaUKMA Research Papers. Computer Science 4 (December 10, 2021): 101–7. http://dx.doi.org/10.18523/2617-3808.2021.4.101-107.
Full textAbstract:
Communication networks are complex information systems influenced by a vast amount of factors. It is critically important to forecast the paths that data take to verify the network, check its security and plan its updates. Model allows exploring processes that take place in the network without affecting performance and availability of a real network itself. With modelling it becomes possible to investigate the results of infrastructural changes introduced to the network before actually implementing them. It is important to be able to formally convert real network description into the model definition which preserves all data that is significant for network operation and skip data which is not. Outlining the rules for such conversion and using a limited set of basic functional components provide the ground for automatic model creation for the network of different levels of complexity.Proposed approach to modelling of communication networks is based on decomposition of the overall function of every particular real network component into a set of functions that belong to some predefined basic set. Functions of the basic set include L3 routing, L2 switching, packet filtering, NAT, etc. Model of a real network component is defined as a group of functional nodes each of which implements some function from the basic set.Configuration and current state of network components that influence its operation are also decomposed into elements each of which relates to some particular functional node. Configuration of network components is modelled as a set of configuration storage elements and current state is modelled as a set of current state storage elements.Links that connect real network components and links that connect functional nodes in the model are presented as singledirection channels that implement propagation of L2 frames thus simplifying the model due to excluding physical layer (L1) from the scope.Using the proposed approach to modelling may allow to formalize conversion of a real network descrip- tion to a model thus making automated modelling possible. By using a sufficient basic set of functional nodes it is possible to model the network containing components of any complexity level.
20
Jesus, Ricardo J., Mário L. Antunes, Rui A. da Costa, Sergey N. Dorogovtsev, José F. F. Mendes, and Rui L. Aguiar. "Effect of Initial Configuration of Weights on Training and Function of Artificial Neural Networks." Mathematics 9, no. 18 (September 13, 2021): 2246. http://dx.doi.org/10.3390/math9182246.
Full textAbstract:
The function and performance of neural networks are largely determined by the evolution of their weights and biases in the process of training, starting from the initial configuration of these parameters to one of the local minima of the loss function. We perform the quantitative statistical characterization of the deviation of the weights of two-hidden-layer feedforward ReLU networks of various sizes trained via Stochastic Gradient Descent (SGD) from their initial random configuration. We compare the evolution of the distribution function of this deviation with the evolution of the loss during training. We observed that successful training via SGD leaves the network in the close neighborhood of the initial configuration of its weights. For each initial weight of a link we measured the distribution function of the deviation from this value after training and found how the moments of this distribution and its peak depend on the initial weight. We explored the evolution of these deviations during training and observed an abrupt increase within the overfitting region. This jump occurs simultaneously with a similarly abrupt increase recorded in the evolution of the loss function. Our results suggest that SGD’s ability to efficiently find local minima is restricted to the vicinity of the random initial configuration of weights.
21
Zhang, Sheng Wen, and Xuan Peng Wang. "Configuration of Multi-Tenant Applications." Advanced Materials Research 219-220 (March 2011): 1182–85. http://dx.doi.org/10.4028/www.scientific.net/amr.219-220.1182.
Full textAbstract:
SaaS (Software as a Service) application mode came into being along with the expansion and convenient of the network platform, and its core lies in Multi-Tenant (multi-tenant) application. Multi-Tenant Application under the SaaS model makes software applications efficient and convenient, more importantly, tenants can greatly reduce their software development costs, hardware acquisition costs, training costs, and upgrade and maintenance costs by using the application system, which virtually eased business cost pressures and makes more focus on business development. This paper elaborates the personalized needs of tenants for Multi-Tenant Applications mainly from the three aspects: data, function, and interface. With the configuration of data, function and interface, Multi-tenant applications will become better.
22
Yotov, Kostadin, Emil Hadzhikolev, Stanka Hadzhikoleva, and Stoyan Cheresharov. "Finding the Optimal Topology of an Approximating Neural Network." Mathematics 11, no. 1 (January 1, 2023): 217. http://dx.doi.org/10.3390/math11010217.
Full textAbstract:
A large number of researchers spend a lot of time searching for the most efficient neural network to solve a given problem. The procedure of configuration, training, testing, and comparison for expected performance is applied to each experimental neural network. The configuration parameters—training methods, transfer functions, number of hidden layers, number of neurons, number of epochs, and tolerable error—have multiple possible values. Setting guidelines for appropriate parameter values would shorten the time required to create an efficient neural network, facilitate researchers, and provide a tool to improve the performance of automated neural network search methods. The task considered in this paper is related to the determination of upper bounds for the number of hidden layers and the number of neurons in them for approximating artificial neural networks trained with algorithms using the Jacobi matrix in the error function. The derived formulas for the upper limits of the number of hidden layers and the number of neurons in them are proved theoretically, and the presented experiments confirm their validity. They show that the search for an efficient neural network can focus below certain upper bounds, and above them, it becomes pointless. The formulas provide researchers with a useful auxiliary tool in the search for efficient neural networks with optimal topology. They are applicable to neural networks trained with methods such as Levenberg–Marquardt, Gauss–Newton, Bayesian regularization, scaled conjugate gradient, BFGS quasi-Newton, etc., which use the Jacobi matrix.
23
Seo, Y., and S. Y. Park. "Prediction of direct runoff hydrographs utilizing stochastic network models: a case study in South Korea." Hydrology and Earth System Sciences Discussions 11, no. 10 (October 10, 2014): 11247–79. http://dx.doi.org/10.5194/hessd-11-11247-2014.
Full textAbstract:
Abstract. In this study, we combine stochastic network models that reproduce the actual width function and the width function based instantaneous unit hydrograph (WFIUH) that directly makes use of a width function and converts it into runoff hydrographs. We evaluated the stochastic network models in terms of reproducing the actual width function and also the robustness of the semi-distributed model (WFIUH) in application to a test watershed in South Korea. The stochastic network model has an advantage that it replicates width functions of actual river networks, whereas the WFIUH has an advantage that the parameter values are physically determined, which can be potentially advantageous in prediction of ungauged basins. This study demonstrates that the combination of the Gibbsian model and the WFIUH is able to reproduce runoff hydrographs not just for the case of uniform rainfall over the test catchment but also for moving storms. Therefore, results of this study indicate that the impact of spatial and temporal rainfall variation on runoff hydrographs can be evaluated by the suggested approach in ungauged basins even without detailed knowledge of river networks. Once the regional similarity in river network configuration is identified, the proposed approach can be potentially utilized to estimate the runoff hydrographs for ungauged basins.
24
Al-Qutt, Mirvat Mahmoud, Heba Khaled, and Rania El Gohary. "Neural Network Inversion-Based Model for Predicting an Optimal Hardware Configuration." International Journal of Grid and High Performance Computing 13, no. 2 (April 2021): 95–117. http://dx.doi.org/10.4018/ijghpc.2021040106.
Full textAbstract:
Deciding the number of processors that can efficiently speed-up solving a computationally intensive problem while perceiving efficient power consumption constitutes a major challenge to researcher in the HPC high performance computing realm. This paper exploits machine learning techniques to propose and implement a recommender system that recommends the optimal HPC architecture given the problem size. An approach for multi-objective function optimization based on neural network (neural network inversion) is employed. The neural network inversion approach is used for forward problem optimization. The objective functions in concern are maximizing the speedup and minimizing the power consumption. The recommendations of the proposed prediction systems achieved more than 89% accuracy for both validation and testing set. The experiments were conducted on 2500 CUDA core on Tesla K20 Kepler GPU Accelerator and Intel(R) Xeon(R) CPU E5-2695 v2.
25
LeCluyse, E. L., K. L. Audus, and J. H. Hochman. "Formation of extensive canalicular networks by rat hepatocytes cultured in collagen-sandwich configuration." American Journal of Physiology-Cell Physiology 266, no. 6 (June 1, 1994): C1764—C1774. http://dx.doi.org/10.1152/ajpcell.1994.266.6.c1764.
Full textAbstract:
Rat primary hepatocytes were cultured under different extracellular matrix configurations and evaluated for the acquisition and maintenance of structural and functional cell polarity. De novo repolarization of the plasma membrane was variable in rate and extent in hepatocyte cultures maintained on a conventional single layer of either gelled or ungelled collagen. However, cultures maintained in a collagen-sandwich configuration initiated uniform formation of a contiguous anastomosing network of bile canaliculi throughout the entire culture. Localization of apical membrane markers demonstrated normal distribution at the canalicular membrane. A marked rearrangement of the intracellular microfilaments to the cell periphery was observed and coincided with the development of the bile canaliculi. Acquisition of normal bile canalicular function and integrity was observed within 3-4 days postoverlay as indicated by the concentration and retention of carboxyfluorescein within the canalicular network. These results demonstrate that cultures of hepatocytes maintained in a sandwich configuration may serve as a more reliable and representative model in which to study the physiology of hepatic function as well as the morphogenesis of polarized membrane domains in vitro.
26
Dou, Zhenlan, Chunyan Zhang, Renjie Dai, Siming Wei, Jihang Zhang, Lingling Wang, and Chuanwen Jiang. "Resilience enhancement strategy of multi-energy coupling distribution network considering movable energy storage equipment." Journal of Physics: Conference Series 2418, no. 1 (February 1, 2023): 012070. http://dx.doi.org/10.1088/1742-6596/2418/1/012070.
Full textAbstract:
Abstract With the deepening of energy transformation, multi-energy coupling has become an effective method to alleviate the energy gap and improve the energy utilization rate. As a bridge connecting users and the power system, the distribution network is becoming multi-energy coupling as well. Therefore, to ensure a stable supply of power, heat, and other energy sources, it is of great significance to study the resilience improvement of multi-energy coupling distribution networks. This paper proposes a strategy for improving the resilience of multi-energy coupling distribution networks considering movable energy storage equipment (MESS) configuration. Firstly, the multi-energy coupling distribution network model is established, and the mathematical model of each piece of equipment is established considering the coupling of the power grid and heat grid. Secondly, based on the loss of load probability (LOLP), considering the importance of the load, it is expanded to the loss of important load probability as an index to evaluate the resilience of distribution networks. Thirdly, a distribution network resilience improvement model based on the configuration and operation of the MESS is established, and the optimal configuration capacity and operation strategy of the MESS is solved by taking the minimum operation cost as the objective function, i.e., the configuration cost of the MESS, the load shedding cost and the cost of wind and PV abandonment. Finally, based on the IEEE-33 node distribution network, a multi-energy coupling distribution network model is built, which verifies that the method proposed in this paper can effectively reduce the total cost and improve the resilience of the multi-energy coupling distribution network.
27
Tang, Yongbin, and Xi Chen. "Software Development, Configuration, Monitoring, and Management of Artificial Neural Networks." Security and Communication Networks 2022 (April 14, 2022): 1–11. http://dx.doi.org/10.1155/2022/9122908.
Full textAbstract:
With the increasing demand for software systems, the software development industry is also developing rapidly. With the development of information technology, the more functions of the software, the more valuable it is, so the function design of the software becomes more complicated and difficult. The design of software system functions is increasingly large and complex. Scientific and effective use of software configuration management can well deal with collaborative work problems such as version management and change control in the software development process. In the process of software development and configuration, there will always be many problems that are difficult to detect. For example, when inputting the program code, there are not always some letter or space errors, and these errors are difficult to detect in time. For this reason, we need to establish a monitoring and management system for software development. As a computing model of human brain neural network, the artificial neural network can play the role of monitoring and management when it is applied to software development and configuration, which provides support for the security and scientificity of software development and configuration systems. This study studies the role and effectiveness of an artificial neural network in the monitoring and management of software development and configuration and validates it through experiments. The experimental results show that the artificial neural network has a strong ability to identify the problems in the software development configuration, which can improve the software development efficiency by at least 20%. It can improve the quality of software development and then improve the life cycle of software.
28
TANAKA, GOUHEI, and KAZUYUKI AIHARA. "MULTISTATE ASSOCIATIVE MEMORY WITH PARAMETRICALLY COUPLED MAP NETWORKS." International Journal of Bifurcation and Chaos 15, no. 04 (April 2005): 1395–410. http://dx.doi.org/10.1142/s0218127405012673.
Full textAbstract:
The present paper proposes two types of parametrically coupled circle map networks for multistate associative memory. One of the networks uses a circle map exhibiting an attractor-merging crisis of multiple chaotic attractors to represent a multistate element. The other uses another circle map whose bifurcation diagram serves as a substitute for a multilevel activation function. The configuration of each network is suitably selected according to the dynamics of the individual circle map so that the network can bring about self-organizing chaotic dynamics with an association of a memory. Namely, the coupling term is determined by the generalized partial error function in the first network, and by the weighted sum of inputs in the second network. These multistate networks can be considered as extensions of two kinds of interesting binary networks called the parametrically coupled sine map networks [Lee & Farhat, 2001a], respectively. We illustrate that the proposed networks can exhibit desirable associative dynamics that is missing in the conventional multistate networks.
29
LI, KANG, and JIAN-XUN PENG. "SYSTEM ORIENTED NEURAL NETWORKS — PROBLEM FORMULATION, METHODOLOGY AND APPLICATION." International Journal of Pattern Recognition and Artificial Intelligence 20, no. 02 (March 2006): 143–58. http://dx.doi.org/10.1142/s0218001406004570.
Full textAbstract:
A novel methodology is proposed for the development of neural network models for complex engineering systems exhibiting nonlinearity. This method performs neural network modeling by first establishing some fundamental nonlinear functions from a priori engineering knowledge, which are then constructed and coded into appropriate chromosome representations. Given a suitable fitness function, using evolutionary approaches such as genetic algorithms, a population of chromosomes evolves for a certain number of generations to finally produce a neural network model best fitting the system data. The objective is to improve the transparency of the neural networks, i.e. to produce physically meaningful "white box" neural network model with better generalization performance. In this paper, the problem formulation, the neural network configuration, and the associated optimization software are discussed in detail. This methodology is then applied to a practical real-world system to illustrate its effectiveness.
30
Laudani, Antonino, Gabriele Maria Lozito, Francesco Riganti Fulginei, and Alessandro Salvini. "On Training Efficiency and Computational Costs of a Feed Forward Neural Network: A Review." Computational Intelligence and Neuroscience 2015 (2015): 1–13. http://dx.doi.org/10.1155/2015/818243.
Full textAbstract:
A comprehensive review on the problem of choosing a suitable activation function for the hidden layer of a feed forward neural network has been widely investigated. Since the nonlinear component of a neural network is the main contributor to the network mapping capabilities, the different choices that may lead to enhanced performances, in terms of training, generalization, or computational costs, are analyzed, both in general-purpose and in embedded computing environments. Finally, a strategy to convert a network configuration between different activation functions without altering the network mapping capabilities will be presented.
31
Bendaoud, Fayssal, Marwen Abdennebi, and Fedoua Didi. "Mobility Aware Network Selection in a Heterogeneous Wireless Environment." Transport and Telecommunication Journal 21, no. 1 (February 1, 2020): 32–46. http://dx.doi.org/10.2478/ttj-2020-0003.
Full textAbstract:
AbstractDifferent Radio Access Technologies (RATs) coexist in the same area has encouraged the researchers to get profit from the available networks by selecting of the best RAT at each moment of the call session to satisfy the user requirements. In this paper, we address a real-world problem which is the frequent mobility of the users in heterogeneous networks. We present in this paper a framework that allows users to select the best networks for the whole call session especially form a mobility perspective. The framework consists of several steps, starting by the path prediction which is performed using a Markov model order 2. The second step is to make the network selection on the zones of each predicted path, while in the third step; we get the best RAT’s configuration for each predicted path. Finally, we use another function to select one of the best configurations to be used for all the possible used paths. The results show that our proposal performs very well by eliminating the unnecessary vertical handover while maintaining a good Quality of Service (QoS).
32
Liu, S. H., Z. J. Bi, and W. Zhang. "A Radar Fault Prediction Based on LM-BP Neural Network." Applied Mechanics and Materials 241-244 (December 2012): 293–97. http://dx.doi.org/10.4028/www.scientific.net/amm.241-244.293.
Full textAbstract:
In order to overcome the difficulty of modern radar fault prediction，which induced by the complexity of system compose，fuzziness of configuration connection and incompletely and uncertainty of character parameters，a LM-BP neural networ model is studied based on BP neural networ model and LM optimized algorithm to optimize the network error function and increase the prediction precision of this model. The simulation and analysis are finished using a radar fault prediction example and show validity the of this model.
33
Meziane, Rachid, Habib Hamdaoui, Mustapha Rahli, and Abdelkader Zeblah. "Structure optimization of electrical power network using ant colony approach." Facta universitatis - series: Electronics and Energetics 16, no. 2 (2003): 233–50. http://dx.doi.org/10.2298/fuee0302233m.
Full textAbstract:
This paper describes and uses an ant colony meta-heuristic optimization method to solve the redundancy optimization problem. This problem is known as total investment-cost minimization of series-parallel power system configuration. Redundant components are included to achieve a desired level of availability. System availability is represented by a multi-state availability function. The power systems components are characterized by their performance (capacity), availability and cost. These components are chosen among a list of products available on the market. The proposed meta-heuristic seeks to the best minimal cost power system configuration with desired availability. To estimate the series-parallel power system availability, a fast method based on universal moment generating function (UMGF) is suggested. The ant colony approach is used as an optimization technique. An example of electrical power system is presented.
34
Yuan, Shuai, Qian Qi, Enliang Dai, and Yongfeng Liang. "Human Resource Planning and Configuration Based on Machine Learning." Computational Intelligence and Neuroscience 2022 (March 15, 2022): 1–6. http://dx.doi.org/10.1155/2022/3605722.
Full textAbstract:
Human resources are the core resources of an enterprise, and the demand forecasting plays a vital role in the allocation and optimization of human resources. Starting from the basic concepts of human resource forecasting, this paper employs the backpropagation neural network (BPNN) and radial basis function neural network (RBFNN) to analyze human resource needs and determine the key elements of the company’s human resource allocation through predictive models. With historical data as reference, the forecast value of current human resource demand is obtained through the two types of neural networks. Based on the prediction results, the company managers can carry out targeted human resource planning and allocation to improve the efficiency of enterprise operations. In the experiment, the actual human resource data of a certain company are used as the experimental basic samples to train and test the two types of machine learning tools. The experimental results show that the method proposed in this paper can effectively predict the number of personnel required and can support the planning and allocation of human resources.
35
Ameen, Shivan Qasim, and Ravie Chandren Muniyandi. "Improvement at Network Planning using Heuristic Algorithm to Minimize Cost of Distance between Nodes in Wireless Mesh Networks." International Journal of Electrical and Computer Engineering (IJECE) 7, no. 1 (February 1, 2017): 309. http://dx.doi.org/10.11591/ijece.v7i1.pp309-315.
Full textAbstract:
Wireless Mesh Networks (WMN) consists of wireless stations that are connected with each other in a semi-static configuration. Depending on the configuration of a WMN, different paths between nodes offer different levels of efficiency. One areas of research with regard to WMN is cost minimization. A Modified Binary Particle Swarm Optimization (MBPSO) approach was used to optimize cost. However, minimized cost does not guarantee network performance. This paper thus, modified the minimization function to take into consideration the distance between the different nodes so as to enable better performance while maintaining cost balance. The results were positive with the PDR showing an approximate increase of 17.83% whereas the E2E delay saw an approximate decrease of 8.33%.
36
Xia, Chuan Kun, Peng Fei Yang, and Si Qing Liu. "Applications of the Optimal Configuration of Section Switches in the Digital System of the Distribution Network Wiring Diagram." Applied Mechanics and Materials 643 (September 2014): 368–73. http://dx.doi.org/10.4028/www.scientific.net/amm.643.368.
Full textAbstract:
the optimal configuration of section switches requires the nonlinear integer programming and the technology, reliability and economy should be taken into account. Therefore, this paper put forward a planning method to determine the optimal additional number and installation position of section switches, which is based on the original frame and switches and combined with the drawing functions in the digital system of the distribution network wiring diagram. The objective function of this method considers the annual investment cost of section switches, the annual operation maintenance cost and the annual loss in outage. It also gives consideration to the correction of investment constraints on the jurisdiction index interval to calculate the number of required additional section switches. Finally, the digital system of the distribution network wiring diagram is utilized to simulate the actual configuration of switches, and the configuration of section switches is optimized by simulating faults through disconnecting the power supply.
37
Gozali, H. R. B. Moch, Suprihadi Prasetyono, and Dimas Oktasya Eka Kumala Putra. "ANALISIS PERBANDINGAN KEANDALAN SISTEM JARINGAN DISTRIBUSI BERKONFIGURASI RADIAL DAN LOOP MENGGUNAKAN METODE RIA (RELIABILITY INDEX ASSESSMENT)." Jurnal Arus Elektro Indonesia 6, no. 3 (December 31, 2020): 63. http://dx.doi.org/10.19184/jaei.v6i3.19723.
Full textAbstract:
Electrical energy is a very important requirement for human life. The supply of electrical energy itself consists of electricity generation, power transmission and distribution. In the distribution of electricity, there are several networks that function to distribute electrical energy to customers, the majority of which are network configuration, namely radial and loop. In order to maintain the continuity of the distribution of electrical energy, a good level of reliability is required. In order to determine the reliability of the distribution system, the reliability index is determined, namely SAIFI, SAIDI and CAIDI. Several methods that can be used to find a distribution system reliability index include the RIA (Reliability Index Assessment) method, which is an approach used to predict disturbances in a distribution system based on the system topology and data regarding component reliability. In this study, we compared two configurations using one feeder, namely the Giri feeder which is assumed to have a radial configuration and a loop configuration. By using the RIA (Reliability Index Assessment) method on a Giri feeder with a radial configuration, the SAIFI reliability index is obtained at 1.595 times / year, SAIDI is 12.8092 hours / year, CAIDI is 8.0309 hours / year, while for a feeder with a loop configuration the SAIFI value is equal to 1,651 times / year, SAIDI is 1,7276 hours / year, CAIDI is 1,0416 hours / year.
Keywords — Reliability Index, Distribution System, Reliability Index Assessment, SAIFI, SAIDI, CAIDI
38
Aghazadeh Ayoubi, Reza, and Umberto Spagnolini. "Performance of Dense Wireless Networks in 5G and beyond Using Stochastic Geometry." Mathematics 10, no. 7 (April 2, 2022): 1156. http://dx.doi.org/10.3390/math10071156.
Full textAbstract:
Device density in cellular networks is expected to increase considerably in the near future. Accordingly, the access point (AP) will be equipped with massive multiple-input multiple-output (mMIMO) antennas, using collimated millimeter-wave (mmW) and sub-THz communications, and increasing the bandwidth to accommodate the growing data rate demands. In this scenario, interference plays a critical role and, if not characterized and mitigated properly, might limit the performances of the network. In this context, this paper derives the statistical properties of the aggregated interference power for a cellular network equipping a mMIMO cylindrical array. The proposed statistical model considers the link blockage and other network parameters such as antenna configuration and device density. The findings show that the characteristic function (CF) of the aggregated interference power can be regarded as a weighted mixture of two alpha-stable distributions. Furthermore, by analyzing the service probability, it is found that there is an optimal configuration of the array depending on the AP height and device density. The proposed statistical model can be part of the design of dense networks providing valuable insights for optimal network deployment and resource management and scheduling.
39
Santhanam, Ganesh Ram, Yuly Suvorov, Samik Basu, and Vasant Honavar. "Verifying Intervention Policies to Counter Infection Propagation over Networks: A Model Checking Approach." Proceedings of the AAAI Conference on Artificial Intelligence 25, no. 1 (August 4, 2011): 1408–14. http://dx.doi.org/10.1609/aaai.v25i1.7804.
Full textAbstract:
Spread of infections (diseases, ideas, etc.) in a networkcan be modeled as the evolution of states of nodes ina graph as a function of the states of their neighbors.Given an initial configuration of a network in which asubset of the nodes have been infected, and an infectionpropagation function that specifies how the states ofthe nodes evolve over time, we show how to use modelchecking to identify, verify, and evaluate the effectivenessof intervention policies for containing the propagationof infection over such networks.
40
Niu, Haoyu, Jiamin Wei, and YangQuan Chen. "Optimal Randomness for Stochastic Configuration Network (SCN) with Heavy-Tailed Distributions." Entropy 23, no. 1 (December 31, 2020): 56. http://dx.doi.org/10.3390/e23010056.
Full textAbstract:
Stochastic Configuration Network (SCN) has a powerful capability for regression and classification analysis. Traditionally, it is quite challenging to correctly determine an appropriate architecture for a neural network so that the trained model can achieve excellent performance for both learning and generalization. Compared with the known randomized learning algorithms for single hidden layer feed-forward neural networks, such as Randomized Radial Basis Function (RBF) Networks and Random Vector Functional-link (RVFL), the SCN randomly assigns the input weights and biases of the hidden nodes in a supervisory mechanism. Since the parameters in the hidden layers are randomly generated in uniform distribution, hypothetically, there is optimal randomness. Heavy-tailed distribution has shown optimal randomness in an unknown environment for finding some targets. Therefore, in this research, the authors used heavy-tailed distributions to randomly initialize weights and biases to see if the new SCN models can achieve better performance than the original SCN. Heavy-tailed distributions, such as Lévy distribution, Cauchy distribution, and Weibull distribution, have been used. Since some mixed distributions show heavy-tailed properties, the mixed Gaussian and Laplace distributions were also studied in this research work. Experimental results showed improved performance for SCN with heavy-tailed distributions. For the regression model, SCN-Lévy, SCN-Mixture, SCN-Cauchy, and SCN-Weibull used less hidden nodes to achieve similar performance with SCN. For the classification model, SCN-Mixture, SCN-Lévy, and SCN-Cauchy have higher test accuracy of 91.5%, 91.7% and 92.4%, respectively. Both are higher than the test accuracy of the original SCN.
41
Guo, Qi, Man Li, Weilin Wang, and Ying Liu. "A Dynamic Deployment Method of Security Services Based on Malicious Behavior Knowledge Base." Sensors 22, no. 22 (November 21, 2022): 9021. http://dx.doi.org/10.3390/s22229021.
Full textAbstract:
In view of various security requirements, there are various security services in the network. In particular, DDoS attacks have various types and detection methods. How to flexibly combine security services and make full use of the information provided by security services have become urgent problems to be solved. This paper combines the reasoning ability of the malicious behavior knowledge base to realize the dynamic deployment of the service function chain and dynamic configuration of the security service function. The method feeds back the information generated by the security service to the knowledge base. After the analysis of the knowledge base, the service function chain path and the security service configuration policies are generated, and these policies will be dynamically distributed to the security service function. Finally, security services can be dynamically arranged for different network traffic, realizing the coordinated use of various security services and improving the overall detection rate of the network. The experimental results show that by arranging the paths under the UDP and the TCP, the overall detection rate of the network can reach 99% and 88%, respectively, indicating that it has a good overall detection performance for multiple distributed denial of service (DDoS) attacks.
42
Zheng, Guang L., and Steve A. Billings. "Radial Basis Function Network Configuration Using Mutual Information and the Orthogonal Least Squares Algorithm." Neural Networks 9, no. 9 (December 1996): 1619–37. http://dx.doi.org/10.1016/0893-6080(95)00139-5.
Full text
43
Dureja, Aman, and Payal Pahwa. "Analysis of Non-Linear Activation Functions for Classification Tasks Using Convolutional Neural Networks." Recent Patents on Computer Science 12, no. 3 (May 8, 2019): 156–61. http://dx.doi.org/10.2174/2213275911666181025143029.
Full textAbstract:
Background: In making the deep neural network, activation functions play an important role. But the choice of activation functions also affects the network in term of optimization and to retrieve the better results. Several activation functions have been introduced in machine learning for many practical applications. But which activation function should use at hidden layer of deep neural networks was not identified. Objective: The primary objective of this analysis was to describe which activation function must be used at hidden layers for deep neural networks to solve complex non-linear problems. Methods: The configuration for this comparative model was used by using the datasets of 2 classes (Cat/Dog). The number of Convolutional layer used in this network was 3 and the pooling layer was also introduced after each layer of CNN layer. The total of the dataset was divided into the two parts. The first 8000 images were mainly used for training the network and the next 2000 images were used for testing the network. Results: The experimental comparison was done by analyzing the network by taking different activation functions on each layer of CNN network. The validation error and accuracy on Cat/Dog dataset were analyzed using activation functions (ReLU, Tanh, Selu, PRelu, Elu) at number of hidden layers. Overall the Relu gave best performance with the validation loss at 25th Epoch 0.3912 and validation accuracy at 25th Epoch 0.8320. Conclusion: It is found that a CNN model with ReLU hidden layers (3 hidden layers here) gives best results and improve overall performance better in term of accuracy and speed. These advantages of ReLU in CNN at number of hidden layers are helpful to effectively and fast retrieval of images from the databases.
44
SHANG, YILUN. "DISTRIBUTION DYNAMICS FOR SIS MODEL ON RANDOM NETWORKS." Journal of Biological Systems 20, no. 02 (June 2012): 213–20. http://dx.doi.org/10.1142/s0218339012500076.
Full textAbstract:
We study the evolution of degree distributions of susceptible-infected-susceptible (SIS) model on random networks, where susceptible nodes are capable of being infected, and infected nodes can spread the disease further. The network of contacts is modeled as a configuration model featuring heterogeneous degree distribution. We derive systematically the (excess) degree distributions among susceptible and infected individuals by using the probability generating function formalism.
45
Oh, Sangchul, Abdelkader Baggag, and Hyunchul Nha. "Entropy, Free Energy, and Work of Restricted Boltzmann Machines." Entropy 22, no. 5 (May 11, 2020): 538. http://dx.doi.org/10.3390/e22050538.
Full textAbstract:
A restricted Boltzmann machine is a generative probabilistic graphic network. A probability of finding the network in a certain configuration is given by the Boltzmann distribution. Given training data, its learning is done by optimizing the parameters of the energy function of the network. In this paper, we analyze the training process of the restricted Boltzmann machine in the context of statistical physics. As an illustration, for small size bar-and-stripe patterns, we calculate thermodynamic quantities such as entropy, free energy, and internal energy as a function of the training epoch. We demonstrate the growth of the correlation between the visible and hidden layers via the subadditivity of entropies as the training proceeds. Using the Monte-Carlo simulation of trajectories of the visible and hidden vectors in the configuration space, we also calculate the distribution of the work done on the restricted Boltzmann machine by switching the parameters of the energy function. We discuss the Jarzynski equality which connects the path average of the exponential function of the work and the difference in free energies before and after training.
46
Nguyen, Thuan Thanh. "Optimization of distribution network configuration with multi objective function based on improved cuckoo search algorithm." Bulletin of Electrical Engineering and Informatics 9, no. 4 (August 1, 2020): 1685–93. http://dx.doi.org/10.11591/eei.v9i4.1886.
Full textAbstract:
This paper proposes an improved cuckoo search (ICSA) for solving the distribution network reconfiguration (NR) problem with multi-objective function. The membership functions are considered consisting of minimizing of power loss, load balancing among branches and among the feeders, node voltage deviation and switching operation numbers. ICSA is developed from the original CSA with adding the local search mechanism for exploiting around the current best solution. The effectiveness of the ICSA has validated on the 70-node and the 83-node practical systems. The obtained results have been compared to those from runner root algorithm (RRA) and other methods in the literature. The obtained results demonstrate that ICSA has high ability for searching the optimal solution with higher successful rate and better quality of obtained solution as well as smaller iterations compared to RRA and other methods. Therefore, ICSA is a reliable method for the multi-objective NR problems.
47
Meqdadi, Osama, Thomas E. Johnsen, and Mark Pagell. "Relationship configurations for procuring from social enterprises." International Journal of Operations & Production Management 40, no. 6 (June 26, 2020): 819–45. http://dx.doi.org/10.1108/ijopm-07-2019-0523.
Full textAbstract:
PurposeThis paper explores how the procurement function initiates and develops relationships with social enterprises that are intended to induce social impact in the supply networks of for-profit firms.Design/methodology/approachThe paper utilises an in-depth case study involving a focal company, first-tier supplier, nongovernmental organisation and four social enterprises.FindingsTension mitigation that arises between social and commercial logics occurs via individual relationships through building trust, dependency manipulation, monitoring and supplier development activities. Deeper insights are revealed when triadic relationships are viewed within a quadratic relationship configuration that enables better capturing the essence of supply networks.Research limitations/implicationsThe paper is based on a single case study, limiting empirical generalisability. Future research could consider multiple case studies to reveal different types of relationship configurations that induce social impact in supply networks.Practical implicationsSocietal goals can be met while maintaining supply network economic performance if procurement involves a trusted third party such as a nongovernmental organisation and helps to develop social enterprises as suppliers.Originality/valueThe paper contributes to the sustainable supply chain management literature by reporting on a novel procurement approach for enhancing social sustainability through cooperation with social enterprises. The paper also contributes to supply network theory by demonstrating how exploring quadratic relationships can reveal novel relationship configurations within supply networks.
48
Kocher, Idrees Sarhan. "AN EXPERIMENTAL SIMULATION OF ADDRESSING AUTO-CONFIGURATION ISSUES FOR WIRELESS SENSOR NETWORKS." Academic Journal of Nawroz University 10, no. 4 (November 25, 2021): 1–13. http://dx.doi.org/10.25007/ajnu.v10n4a957.
Full textAbstract:
Wireless Sensor Networks (WSN) is now an evolving technology and has a broad range of applications, such as battlefield surveillance, traffic surveillance, detection of forest fires, detection of floods, etc. The communication nature of the wireless sensor network is unprotected and dangerous due to deployment in hostile environments, restricted resources, an automatic nature, and untrusted media for broadcast transmission. For wireless sensor networks, several routing protocols have been suggested, but none of them have been developed with protection as a target. The majority function in routing algorithms currently in place for sensor networks optimize a restricted capacities in sensor nodes and the application based design of WSNs. A WSNs, however, are exposed to a number of possible threats that impede the network's regular activity. Thus, there is a strong need to provide the routing protocols of the OSI structure layer with a safe mechanism to prevent an attacker from obstructing it.
The well-known attacks against all layers are discussed in this systematic roadmap, and debilitating attacks against routing protocols are analyzed and defined in particular. Several suggested attack countermeasures, design considerations and paper contributions are also included in the routing protocols. The assertion of the study is that WSN routing protocols must be built with protection in mind, and this is the only efficient solution in WSNs for safe routing. The aim of this paper is also to provide problems, attacks and countermeasures related to protection. Finally, it is hoped that this roadmap would inspire potential researchers to come up with smarter and better protection measures and make their network safer.
The first such research analysis of secure routing protocols in WSNs is this roadmap study.
49
Cantoro, Riccardo, Aleksa Damljanovic, Matteo Sonza Reorda, and Giovanni Squillero. "An Enhanced Evolutionary Technique for the Generation of Compact Reconfigurable Scan-Network Tests." Journal of Circuits, Systems and Computers 28, supp01 (December 1, 2019): 1940007. http://dx.doi.org/10.1142/s0218126619400073.
Full textAbstract:
Nowadays, many Integrated Systems embed auxiliary on-chip instruments whose function is to perform test, debug, calibration, configuration, etc. The growing complexity and the increasing number of these instruments have led to new solutions for their access and control, such as the IEEE 1687 standard. The standard introduces an infrastructure composed of scan chains incorporating configurable elements for accessing the instruments in a flexible manner. Such an infrastructure is known as Reconfigurable Scan Network or RSN. Since permanent faults affecting the circuitry can cause malfunction, i.e., inappropriate behavior, detecting them is of utmost importance. This paper addresses the issue of generating effective sequences for testing the reconfigurable elements within RSNs using evolutionary computation. Test configurations are extracted with automatic test pattern generation (ATPG) and used to guide the evolution. Post-processing techniques are proposed to improve the evolutionary fittest solution. Results on a standard set of benchmark networks show up to 27% reduced test time with respect to test generation based on RSN exploration.
50
Bloch, Aurelien, Simone Casale-Brunet, and Marco Mattavelli. "Performance Estimation of High-Level Dataflow Program on Heterogeneous Platforms by Dynamic Network Execution." Journal of Low Power Electronics and Applications 12, no. 3 (June 23, 2022): 36. http://dx.doi.org/10.3390/jlpea12030036.
Full textAbstract:
The performance of programs executed on heterogeneous parallel platforms largely depends on the design choices regarding how to partition the processing on the various different processing units. In other words, it depends on the assumptions and parameters that define the partitioning, mapping, scheduling, and allocation of data exchanges among the various processing elements of the platform executing the program. The advantage of programs written in languages using the dataflow model of computation (MoC) is that executing the program with different configurations and parameter settings does not require rewriting the application software for each configuration setting, but only requires generating a new synthesis of the execution code corresponding to different parameters. The synthesis stage of dataflow programs is usually supported by automatic code generation tools. Another competitive advantage of dataflow software methodologies is that they are well-suited to support designs on heterogeneous parallel systems as they are inherently free of memory access contention issues and naturally expose the available intrinsic parallelism. So as to fully exploit these advantages and to be able to efficiently search the configuration space to find the design points that better satisfy the desired design constraints, it is necessary to develop tools and associated methodologies capable of evaluating the performance of different configurations and to drive the search for good design configurations, according to the desired performance criteria. The number of possible design assumptions and associated parameter settings is usually so large (i.e., the dimensions and size of the design space) that intuition as well as trial and error are clearly unfeasible, inefficient approaches. This paper describes a method for the clock-accurate profiling of software applications developed using the dataflow programming paradigm such as the formal RVL-CAL language. The profiling can be applied when the application program has been compiled and executed on GPU/CPU heterogeneous hardware platforms utilizing two main methodologies, denoted as static and dynamic. This paper also describes how a method for the qualitative evaluation of the performance of such programs as a function of the supplied configuration parameters can be successfully applied to heterogeneous platforms. The technique was illustrated using two different application software examples and several design points.