Search results for: stochastic gradient kernel.

Authors: Paola Lecca, Lorenzo Dematte, Corrado Priami

Abstract:

The present models and simulation algorithms of intracellular stochastic kinetics are usually based on the premise that diffusion is so fast that the concentrations of all the involved species are homogeneous in space. However, recents experimental measurements of intracellular diffusion constants indicate that the assumption of a homogeneous well-stirred cytosol is not necessarily valid even for small prokaryotic cells. In this work a mathematical treatment of diffusion that can be incorporated in a stochastic algorithm simulating the dynamics of a reaction-diffusion system is presented. The movement of a molecule A from a region i to a region j of the space is represented as a first order reaction Ai k- ! Aj , where the rate constant k depends on the diffusion coefficient. The diffusion coefficients are modeled as function of the local concentration of the solutes, their intrinsic viscosities, their frictional coefficients and the temperature of the system. The stochastic time evolution of the system is given by the occurrence of diffusion events and chemical reaction events. At each time step an event (reaction or diffusion) is selected from a probability distribution of waiting times determined by the intrinsic reaction kinetics and diffusion dynamics. To demonstrate the method the simulation results of the reaction-diffusion system of chaperoneassisted protein folding in cytoplasm are shown.

Keywords: Reaction-diffusion systems, diffusion coefficient, stochastic simulation algorithm.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1482

Authors: Romisaa Ali

Abstract:

The Policy Gradient approach is a subset of the Deep Reinforcement Learning (DRL) combines Deep Neural Networks (DNN) with Reinforcement Learning (RL). This approach finds the optimal policy of robot movement, based on the experience it gains from interaction with its environment. Unlike previous policy gradient algorithms, which were unable to handle the two types of error variance and bias introduced by the DNN model due to over- or underestimation, this algorithm is capable of handling both types of error variance and bias. This article will discuss the state-of-the-art SOTA policy gradient technique, trust region policy optimization (TRPO), by applying this method in various environments compared to another policy gradient method, the Proximal Policy Optimization (PPO), to explain their robust optimization, using this SOTA to gather experience data during various training phases after observing the impact of hyper-parameters on neural network performance.

Keywords: Deep neural networks, deep reinforcement learning, Proximal Policy Optimization, state-of-the-art, trust region policy optimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 113

Authors: N. C. Sarkar, A. Bhaskar, Z. Zheng

Abstract:

The driving behavior in area-based (i.e., non-lane based) traffic is induced by the presence of other individuals in the choice space from the driver’s visual perception area. The driving behavior of a subject vehicle is constrained by the potential leaders and leaders are frequently changed over time. This paper is to determine a stochastic model for a parameter of modified intelligent driver model (MIDM) in area-based traffic (as in developing countries). The parametric and non-parametric distributions are presented to fit the parameters of MIDM. The goodness of fit for each parameter is measured in two different ways such as graphically and statistically. The quantile-quantile (Q-Q) plot is used for a graphical representation of a theoretical distribution to model a parameter and the Kolmogorov-Smirnov (K-S) test is used for a statistical measure of fitness for a parameter with a theoretical distribution. The distributions are performed on a set of estimated parameters of MIDM. The parameters are estimated on the real vehicle trajectory data from India. The fitness of each parameter with a stochastic model is well represented. The results support the applicability of the proposed modeling for parameters of MIDM in area-based traffic flow simulation.

Keywords: Area-based traffic, car-following model, micro-simulation, stochastic modeling.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 724

Authors: Walid S. Alfuhaid, Saud A. Alghamdi, John M. Watkins, M. Edwin Sawan

Abstract:

Designing a controller for stochastic decentralized interconnected large scale systems usually involves a high degree of complexity and computation ability. Noise, observability, and controllability of all system states, connectivity, and channel bandwidth are other constraints to design procedures for distributed large scale systems. The quasi-steady state model investigated in this paper is a reduced order model of the original system using singular perturbation techniques. This paper results in an optimal control synthesis to design an observer based feedback controller by standard stochastic control theory techniques using Linear Quadratic Gaussian (LQG) approach and Kalman filter design with less complexity and computation requirements. Numerical example is given at the end to demonstrate the efficiency of the proposed method.

Keywords: Decentralized, optimal control, output, singular perturb.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1533

Authors: Abeer Amayri, Akif A. Bulgak

Abstract:

Due to today’s globalization as well as outsourcing practices of the companies, the Supply Chain (SC) performances have become more dependent on the efficient movement of material among places that are geographically dispersed, where there is more chance for disruptions. One such disruption is the quality and delivery uncertainties of outsourcing. These uncertainties could lead the products to be unsafe and, as is the case in a number of recent examples, companies may have to end up in recalling their products. As a result of these problems, there is a need to develop a methodology for selecting suppliers globally in view of risks associated with low quality and late delivery. Accordingly, we developed a two-stage stochastic model that captures the risks associated with uncertainty in quality and delivery as well as a solution procedure for the model. The stochastic model developed simultaneously optimizes supplier selection and purchase quantities under price discounts over a time horizon. In particular, our target is the study of global organizations with multiple sites and multiple overseas suppliers, where the pricing is offered in suppliers’ local currencies. Our proposed methodology is applied to a case study for a US automotive company having two assembly plants and four potential global suppliers to illustrate how the proposed model works in practice.

Keywords: Global supply chains, quality, stochastic programming, supplier selection.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1519

Authors: Nima Farkhondeh Jahromi, George Papaefthymiou, Lou van der Sluis

Abstract:

The incorporation of renewable energy sources for the sustainable electricity production is undertaking a more prominent role in electric power systems. Thus, it will be an indispensable incident that the characteristics of future power networks, their prospective stability for instance, get influenced by the imposed features of sustainable energy sources. One of the distinctive attributes of the sustainable energy sources is exhibiting the stochastic behavior. This paper investigates the impacts of this stochastic behavior on the small disturbance rotor angle stability in the upcoming electric power networks. Considering the various types of renewable energy sources and the vast variety of system configurations, the sensitivity analysis can be an efficient breakthrough towards generalizing the effects of new energy sources on the concept of stability. In this paper, the definition of small disturbance angle stability for future power systems and the iterative-stochastic way of its analysis are presented. Also, the effects of system parameters on this type of stability are described by performing a sensitivity analysis for an electric power test system.

Keywords: Power systems stability, Renewable energy sources, Stochastic behavior, Small disturbance rotor angle stability.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2027

Authors: Mahidin, Yanna Syamsuddin, Samsul Rizal

Abstract:

In order to fulfill world energy demand, several efforts have been done to look for new and renewable energy candidates to substitute oil and gas. Biomass is one of new and renewable energy sources, which is abundant in Indonesia. Palm kernel shell is a kind of biomass discharge from palm oil industries as a waste. On the other hand, Jatropha curcas that is easy to grow in Indonesia is also a typical energy source either for bio-diesel or biomass. In this study, biomass was used as co-fuel in briquetting of low-rank coal to suppress the release of emission (such as CO, NO_x and SO_x) during coal combustion. Desulfurizer, CaO-base, was also added to ensure the SO_x capture is effectively occurred. Ratio of coal to palm kernel shell (w/w) in the bio-briquette were 50:50, 60:40, 70:30, 80:20 and 90:10, while ratio of calcium to sulfur (Ca/S) in mole/mole were 1:1; 1.25:1; 1.5:1; 1.75:1 and 2:1. The bio-briquette then subjected to physical characterization and combustion test. The results show that the maximum weight loss in the durability measurement was ±6%. In addition, the highest stove efficiency for each desulfurizer was observed at the coal/PKS ratio of 90:10 and Ca/S ratio of 1:1 (except for the scallop shell desulfurizer that appeared at two Ca/S ratios; 1.25:1 and 1.5:1, respectively), i.e. 13.8% for the lime; 15.86% for the oyster shell; 14.54% for the scallop shell and 15.84% for the green mussel shell desulfurizers.

Keywords: Biomass, low-rank coal, bio-briquette, new and renewable energy, palm kernel shell.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2612

Authors: A. Elyousfi, A. Tamtaoui, E. Bouyakhf

Abstract:

The H.264/AVC standard uses an intra prediction, 9 directional modes for 4x4 luma blocks and 8x8 luma blocks, 4 directional modes for 16x16 macroblock and 8x8 chroma blocks, respectively. It means that, for a macroblock, it has to perform 736 different RDO calculation before a best RDO modes is determined. With this Multiple intra-mode prediction, intra coding of H.264/AVC offers a considerably higher improvement in coding efficiency compared to other compression standards, but computational complexity is increased significantly. This paper presents a fast intra prediction algorithm for H.264/AVC intra prediction based a characteristic of homogeneity information. In this study, the gradient prediction method used to predict the homogeneous area and the quadratic prediction function used to predict the nonhomogeneous area. Based on the correlation between the homogeneity and block size, the smaller block is predicted by gradient prediction and quadratic prediction, so the bigger block is predicted by gradient prediction. Experimental results are presented to show that the proposed method reduce the complexity by up to 76.07% maintaining the similar PSNR quality with about 1.94%bit rate increase in average.

Keywords: Intra prediction, H.264/AVC, video coding, encodercomplexity.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1839

Authors: Miljan B. Petrović, Dušan B. Petrović, Goran S. Nikolić

Abstract:

This paper describes a method for AWGN (Additive White Gaussian Noise) variance estimation in noisy stochastic signals, referred to as Multiplicative-Noising Variance Estimation (MNVE). The aim was to develop an estimation algorithm with minimal number of assumptions on the original signal structure. The provided MATLAB simulation and results analysis of the method applied on speech signals showed more accuracy than standardized AR (autoregressive) modeling noise estimation technique. In addition, great performance was observed on very low signal-to-noise ratios, which in general represents the worst case scenario for signal denoising methods. High execution time appears to be the only disadvantage of MNVE. After close examination of all the observed features of the proposed algorithm, it was concluded it is worth of exploring and that with some further adjustments and improvements can be enviably powerful.

Keywords: Noise, signal-to-noise ratio, stochastic signals, variance estimation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2211

Authors: Serge B. Provost, Min Jiang

Abstract:

The density estimates considered in this paper comprise a base density and an adjustment component consisting of a linear combination of orthogonal polynomials. It is shown that, in the context of density approximation, the coefficients of the linear combination can be determined either from a moment-matching technique or a weighted least-squares approach. A kernel representation of the corresponding density estimates is obtained. Additionally, two refinements of the Kronmal-Tarter stopping criterion are proposed for determining the degree of the polynomial adjustment. By way of illustration, the density estimation methodology advocated herein is applied to two data sets.

Keywords: kernel density estimation, orthogonal polynomials, moment-based methodologies, density approximation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2322

Authors: Benjamin Gorry, Zezhi Chen, Kevin Hammond, Andy Wallace, Greg Michaelson

Abstract:

This paper describes new computer vision algorithms that have been developed to track moving objects as part of a long-term study into the design of (semi-)autonomous vehicles. We present the results of a study to exploit variable kernels for tracking in video sequences. The basis of our work is the mean shift object-tracking algorithm; for a moving target, it is usual to define a rectangular target window in an initial frame, and then process the data within that window to separate the tracked object from the background by the mean shift segmentation algorithm. Rather than use the standard, Epanechnikov kernel, we have used a kernel weighted by the Chamfer distance transform to improve the accuracy of target representation and localization, minimising the distance between the two distributions in RGB color space using the Bhattacharyya coefficient. Experimental results show the improved tracking capability and versatility of the algorithm in comparison with results using the standard kernel. These algorithms are incorporated as part of a robot test-bed architecture which has been used to demonstrate their effectiveness.

Keywords: Hume, functional programming, autonomous vehicle, pioneer robot, vision.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1610

Authors: Zhifeng Kong

Abstract:

Over-parameterized neural networks have attracted a great deal of attention in recent deep learning theory research, as they challenge the classic perspective of over-fitting when the model has excessive parameters and have gained empirical success in various settings. While a number of theoretical works have been presented to demystify properties of such models, the convergence properties of such models are still far from being thoroughly understood. In this work, we study the convergence properties of training two-hidden-layer partially over-parameterized fully connected networks with the Rectified Linear Unit activation via gradient descent. To our knowledge, this is the first theoretical work to understand convergence properties of deep over-parameterized networks without the equally-wide-hidden-layer assumption and other unrealistic assumptions. We provide a probabilistic lower bound of the widths of hidden layers and proved linear convergence rate of gradient descent. We also conducted experiments on synthetic and real-world datasets to validate our theory.

Keywords: Over-parameterization, Rectified Linear Units (ReLU), convergence, gradient descent, neural networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 812

Authors: Sung-Hae Jun, Kyung-Whan Oh

Abstract:

Statistical learning theory was developed by Vapnik. It is a learning theory based on Vapnik-Chervonenkis dimension. It also has been used in learning models as good analytical tools. In general, a learning theory has had several problems. Some of them are local optima and over-fitting problems. As well, statistical learning theory has same problems because the kernel type, kernel parameters, and regularization constant C are determined subjectively by the art of researchers. So, we propose an evolutionary statistical learning theory to settle the problems of original statistical learning theory. Combining evolutionary computing into statistical learning theory, our theory is constructed. We verify improved performances of an evolutionary statistical learning theory using data sets from KDD cup.

Keywords: Evolutionary computing, Local optima, Over-fitting, Statistical learning theory

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1728

Authors: P. Ninduangdee, V. I. Kuprianov

Abstract:

Palm kernel shell is an important bioenergy resource in Thailand. However, due to elevated alkali content in biomass ash, this oil palm residue shows high tendency to bed agglomeration in a fluidized-bed combustion system using conventional bed material (silica sand). In this study, palm kernel shell was burned in the conical fluidized-bed combustor (FBC) using alumina and dolomite as alternative bed materials to prevent bed agglomeration. For each bed material, the combustion tests were performed at 45kg/h fuel feed rate with excess air within 20–80%. Experimental results revealed rather weak effects of the bed material type but substantial influence of excess air on the behavior of temperature, O2, CO, CxHy, and NO inside the reactor, as well as on the combustion efficiency and major gaseous emissions of the conical FBC. The optimal level of excess air ensuring high combustion efficiency (about 98.5%) and acceptable level of the emissions was found to be about 40% when using alumina and 60% with dolomite. By using these alternative bed materials, bed agglomeration can be prevented when burning the shell in the proposed conical FBC. However, both bed materials exhibited significant changes in their morphological, physical and chemical properties in the course of the time.

Keywords: Palm kernel shell, fluidized-bed combustion, alternative bed materials, combustion and emission performance, bed agglomeration prevention.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2943

Authors: S. Rozhkova, O. Rozhkova, A. Harlova, V. Lasukov

Abstract:

We have conducted the optimal synthesis of rootmean- squared objective filter to estimate the state vector in the case if within the observation channel with memory the anomalous noises with unknown mathematical expectation are complement in the function of the regular noises. The synthesis has been carried out for linear stochastic systems of continuous - time.

Keywords: Mathematical expectation, filtration, anomalous noise, memory.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1924

Authors: Mehdi Ghayoumi

Abstract:

We have applied new accelerated algorithm for linear discriminate analysis (LDA) in face recognition with support vector machine. The new algorithm has the advantage of optimal selection of the step size. The gradient descent method and new algorithm has been implemented in software and evaluated on the Yale face database B. The eigenfaces of these approaches have been used to training a KNN. Recognition rate with new algorithm is compared with gradient.

Keywords: lda, adaptive, svm, face recognition.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1377

Authors: Danni Cong, Meiping Wu, Xiaofeng He, Junxiang Lian, Juliang Cao, Shaokuncai, Hao Qin

Abstract:

Gravity gradient instrument (GGI) is the core of the gravity gradiometer, so the structural error of the sensor has a great impact on the measurement results. In order not to affect the aimed measurement accuracy, limit error is required in the installation of the accelerometer. In this paper, based on the established measuring principle model, the radial installation limit error is calibrated, which is taken as an example to provide a method to calculate the other limit error of the installation under the premise of ensuring the accuracy of the measurement result. This method provides the idea for deriving the limit error of the geometry structure of the sensor, laying the foundation for the mechanical precision design and physical design.

Keywords: Gravity gradient sensor, radial installation limit error, accelerometer, uniaxial rotational modulation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 881

Authors: Iau-Teh Wang, Chin-Yu Lee

Abstract:

The impact force of a rockfall is mainly determined by its moving behavior and velocity, which are contingent on the rock shape, slope gradient, height, and surface roughness of the moving path. It is essential to precisely calculate the moving path of the rockfall in order to effectively minimize and prevent damages caused by the rockfall. By applying the Colorado Rockfall Simulation Program (CRSP) program as the analysis tool, this research studies the influence of three shapes of rock (spherical, cylindrical and discoidal) and surface roughness on the moving path of a single rockfall. As revealed in the analysis, in addition to the slope gradient, the geometry of the falling rock and joint roughness coefficient ( JRC ) of the slope are the main factors affecting the moving behavior of a rockfall. On a single flat slope, both the rock-s bounce height and moving velocity increase as the surface gradient increases, with a critical gradient value of 1:m = 1 . Bouncing behavior and faster moving velocity occur more easily when the rock geometry is more oval. A flat piece tends to cause sliding behavior and is easily influenced by the change of surface undulation. When JRC <1.4 the moving velocity decreases and the bounce height increases as JRC increases. If the gradient is fixed, when JRC is greater, the bounce height will be higher, while the moving velocity will experience a downward trend. Therefore, the best protecting point and facilities can be chosen if the moving paths of rockfalls are precisely estimated.

Keywords: rock shape, surface roughness, moving path.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1907

Authors: M. Shaker, S. Minaei, M. H. Khoshtaghaza, A. Banakar, A. Jafari

Abstract:

In this study a system of machine vision and singulation was developed to separate paddy from rice and determine paddy husking and rice breakage percentages. The machine vision system consists of three main components including an imaging chamber, a digital camera, a computer equipped with image processing software. The singulation device consists of a kernel holding surface, a motor with vacuum fan, and a dimmer. For separation of paddy from rice (in the image), it was necessary to set a threshold. Therefore, some images of paddy and rice were sampled and the RGB values of the images were extracted using MATLAB software. Then mean and standard deviation of the data were determined. An Image processing algorithm was developed using MATLAB to determine paddy/rice separation and rice breakage and paddy husking percentages, using blue to red ratio. Tests showed that, a threshold of 0.75 is suitable for separating paddy from rice kernels. Results from the evaluation of the image processing algorithm showed that the accuracies obtained with the algorithm were 98.36% and 91.81% for paddy husking and rice breakage percentage, respectively. Analysis also showed that a suction of 45 mmHg to 50 mmHg yielding 81.3% separation efficiency is appropriate for operation of the kernel singulation system.

Keywords: Computer vision, rice kernel, husking, breakage.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1471

Authors: Bib Paruhum Silalahi, Djihad Wungguli, Sugi Guritman

Abstract:

Steepest descent method is a simple gradient method for optimization. This method has a slow convergence in heading to the optimal solution, which occurs because of the zigzag form of the steps. Barzilai and Borwein modified this algorithm so that it performs well for problems with large dimensions. Barzilai and Borwein method results have sparked a lot of research on the method of steepest descent, including alternate minimization gradient method and Yuan method. Inspired by previous works, we modified the step size of the steepest descent method. We then compare the modification results against the Barzilai and Borwein method, alternate minimization gradient method and Yuan method for quadratic function cases in terms of the iterations number and the running time. The average results indicate that the steepest descent method with the new step sizes provide good results for small dimensions and able to compete with the results of Barzilai and Borwein method and the alternate minimization gradient method for large dimensions. The new step sizes have faster convergence compared to the other methods, especially for cases with large dimensions.

Keywords: Convergence, iteration, line search, running time, steepest descent, unconstrained optimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3113

Authors: Syed Muhammad Aqil Burney, Tahseen Ahmed Jilani, C. Ardil

Abstract:

Minimization methods for training feed-forward networks with Backpropagation are compared. Feedforward network training is a special case of functional minimization, where no explicit model of the data is assumed. Therefore due to the high dimensionality of the data, linearization of the training problem through use of orthogonal basis functions is not desirable. The focus is functional minimization on any basis. A number of methods based on local gradient and Hessian matrices are discussed. Modifications of many methods of first and second order training methods are considered. Using share rates data, experimentally it is proved that Conjugate gradient and Quasi Newton?s methods outperformed the Gradient Descent methods. In case of the Levenberg-Marquardt algorithm is of special interest in financial forecasting.

Keywords: Backpropagation algorithm, conjugacy condition, line search, matrix perturbation

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3597

Authors: Helle Hein, Ülo Lepik

Abstract:

A gradient learning method to regulate the trajectories of some nonlinear chaotic systems is proposed. The method is motivated by the gradient descent learning algorithms for neural networks. It is based on two systems: dynamic optimization system and system for finding sensitivities. Numerical results of several examples are presented, which convincingly illustrate the efficiency of the method.

Keywords: Chaos, Dynamical Systems, Learning, Neural Networks

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1329

Authors: Mizan Ahmed, Srikanth Venkatesan

Abstract:

Geological and tectonic framework indicates that Bangladesh is one of the most seismically active regions in the world. The Bengal Basin is at the junction of three major interacting plates: the Indian, Eurasian, and Burma Plates. Besides there are many active faults within the region, e.g. the large Dauki fault in the north. The country has experienced a number of destructive earthquakes due to the movement of these active faults. Current seismic provisions of Bangladesh are mostly based on earthquake data prior to the 1990. Given the record of earthquakes post 1990, there is a need to revisit the design provisions of the code. This paper compares the base shear demand of three major cities in Bangladesh: Dhaka (the capital city), Sylhet, and Chittagong for earthquake scenarios of magnitudes 7.0MW, 7.5MW, 8.0MW, and 8.5MW using a stochastic model. In particular, the stochastic model allows the flexibility to input region specific parameters such as shear wave velocity profile (that were developed from Global Crustal Model CRUST2.0) and include the effects of attenuation as individual components. Effects of soil amplification were analysed using the Extended Component Attenuation Model (ECAM). Results show that the estimated base shear demand is higher in comparison with code provisions leading to the suggestion of additional seismic design consideration in the study regions.

Keywords: Attenuation, earthquake, ground motion, stochastic, seismic hazard.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1990

Authors: Daniel Morariu, Lucian N. Vintan, Volker Tresp

Abstract:

Text categorization is the problem of classifying text documents into a set of predefined classes. After a preprocessing step, the documents are typically represented as large sparse vectors. When training classifiers on large collections of documents, both the time and memory restrictions can be quite prohibitive. This justifies the application of feature selection methods to reduce the dimensionality of the document-representation vector. In this paper, three feature selection methods are evaluated: Random Selection, Information Gain (IG) and Support Vector Machine feature selection (called SVM_FS). We show that the best results were obtained with SVM_FS method for a relatively small dimension of the feature vector. Also we present a novel method to better correlate SVM kernel-s parameters (Polynomial or Gaussian kernel).

Keywords: Feature Selection, Learning with Kernels, SupportVector Machine, and Classification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1778

Authors: Myat Su Mon Win

Abstract:

The pseudorandom number generators based on linear feedback shift registers (LFSRs), are very quick, easy and secure in the implementation of hardware and software. Thus they are very popular and widely used. But LFSRs lead to fairly easy cryptanalysis due to their completely linearity properties. In this paper, we propose a stochastic generator, which is called Random Feedback Shift Register (RFSR), using stochastic transformation (Random block) with one-way and non-linearity properties.

Keywords: Linear Feedback Shift Register, Non Linearity, R_Block, Random Feedback Shift Register

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1768

Authors: Weizhi Xu, Zhiyong Liu, Dongrui Fan, Shuai Jiao, Xiaochun Ye, Fenglong Song, Chenggang Yan

Abstract:

Many-core GPUs provide high computing ability and substantial bandwidth; however, optimizing irregular applications like SpMV on GPUs becomes a difficult but meaningful task. In this paper, we propose a novel method to improve the performance of SpMV on GPUs. A new storage format called HYB-R is proposed to exploit GPU architecture more efficiently. The COO portion of the matrix is partitioned recursively into a ELL portion and a COO portion in the process of creating HYB-R format to ensure that there are as many non-zeros as possible in ELL format. The method of partitioning the matrix is an important problem for HYB-R kernel, so we also try to tune the parameters to partition the matrix for higher performance. Experimental results show that our method can get better performance than the fastest kernel (HYB) in NVIDIA-s SpMV library with as high as 17% speedup.

Keywords: GPU, HYB-R, Many-core, Performance Tuning, SpMV

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1949

Authors: Turgay Temel, John Hallam

Abstract:

A recent neurospiking coding scheme for feature extraction from biosonar echoes of various plants is examined with avariety of stochastic classifiers. Feature vectors derived are employedin well-known stochastic classifiers, including nearest-neighborhood,single Gaussian and a Gaussian mixture with EM optimization.Classifiers' performances are evaluated by using cross-validation and bootstrapping techniques. It is shown that the various classifers perform equivalently and that the modified preprocessing configuration yields considerably improved results.

Keywords: Classification, neuro-spike coding, non-parametricmodel, parametric model, Gaussian mixture, EM algorithm.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1623

Authors: S. Chadsuthi, W. Triampo, C. Modchang, P. Kanthang, D. Triampo, N. Nuttavut

Abstract:

We present analysis of spatial patterns of generic disease spread simulated by a stochastic long-range correlation SIR model, where individuals can be infected at long distance in a power law distribution. We integrated various tools, namely perimeter, circularity, fractal dimension, and aggregation index to characterize and investigate spatial pattern formations. Our primary goal was to understand for a given model of interest which tool has an advantage over the other and to what extent. We found that perimeter and circularity give information only for a case of strong correlation– while the fractal dimension and aggregation index exhibit the growth rule of pattern formation, depending on the degree of the correlation exponent (β). The aggregation index method used as an alternative method to describe the degree of pathogenic ratio (α). This study may provide a useful approach to characterize and analyze the pattern formation of epidemic spreading

Keywords: spatial pattern epidemics, aggregation index, fractaldimension, stochastic, long-rang epidemics

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1636

Authors: Xiang Zhang, David Rey, S. Travis Waller

Abstract:

Stochastic User Equilibrium (SUE) model is a widely used traffic assignment model in transportation planning, which is regarded more advanced than Deterministic User Equilibrium (DUE) model. However, a problem exists that the performance of the SUE model depends on its error term parameter. The objective of this paper is to propose a systematic method of determining the appropriate error term parameter value for the SUE model. First, the significance of the parameter is explored through a numerical example. Second, the parameter calibration method is developed based on the Logit-based route choice model. The calibration process is realized through multiple nonlinear regression, using sequential quadratic programming combined with least square method. Finally, case analysis is conducted to demonstrate the application of the calibration process and validate the better performance of the SUE model calibrated by the proposed method compared to the SUE models under other parameter values and the DUE model.

Keywords: Parameter calibration, sequential quadratic programming, Stochastic User Equilibrium, traffic assignment, transportation planning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2078

Authors: Ching-Fen Jiang

Abstract:

Edge detection is usually the first step in medical image processing. However, the difficulty increases when a conventional kernel-based edge detector is applied to ultrasonic images with a textural pattern and speckle noise. We designed an adaptive diffusion filter to remove speckle noise while preserving the initial edges detected by using a Sobel edge detector. We also propose a genetic algorithm for edge selection to form complete boundaries of the detected entities. We designed two fitness functions to evaluate whether a criterion with a complex edge configuration can render a better result than a simple criterion such as the strength of gradient. The edges obtained by using a complex fitness function are thicker and more fragmented than those obtained by using a simple fitness function, suggesting that a complex edge selecting scheme is not necessary for good edge detection in medical ultrasonic images; instead, a proper noise-smoothing filter is the key.

Keywords: edge detection, ultrasonic images, speckle noise

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1441