Search results for: gradient features

Authors: A. A. Abdullah, K. A. Lindsay

Abstract:

The management of a salt-gradient is investigated with respect to the interaction between the solar pond and its associated evaporation pond. Issues considered are the impact of precipitation and the operation of the flushing system with particular reference to the case in which the flushing fluid is pure water. Results suggest that a management strategy based on a flushing system that simply replaces evaporation losses of water from the solar pond and evaporation pond will be optimally efficient. Such a management strategy will maintain the operational viability of a salt-gradient solar pond as a reservoir of cheap heat while simultaneously ensuring that the associated evaporation pond can feed the storage zone of the solar pond with sufficient saturated brine to balance the effect of salt diffusion. Other findings are, first, that once near saturation is achieved in the evaporation pond, the efficacy of the proposed management strategy is relatively insensitive to both the size of the evaporation pond or its depth, and second, small changes in the extraction of heat from the storage zone of a salt-gradient solar pond have an amplified effect on the temperature of that zone. The possibility of boiling of the storage zone cannot be ignored in a well-configured salt-gradient solar pond.

Keywords: aqueous sodium chloride, constitutive expression, solar pond, salt-gradient

Procedia PDF Downloads 303

Authors: Yanwen Li, Shuguo Xie

Abstract:

In electromagnetic imaging, because of the diffraction limited system, the pixel values could change slowly near the edge of the image targets and they also change with the location in the same target. Using traditional digital image segmentation methods to segment electromagnetic gradient images could result in lots of errors because of this change in pixel values. To address this issue, this paper proposes a novel image segmentation and extraction algorithm based on Mean-Shift and dictionary learning. Firstly, the preliminary segmentation results from adaptive bandwidth Mean-Shift algorithm are expanded, merged and extracted. Then the overlap rate of the extracted image block is detected before determining a segmentation region with a single complete target. Last, the gradient edge of the extracted targets is recovered and reconstructed by using a dictionary-learning algorithm, while the final segmentation results are obtained which are very close to the gradient target in the original image. Both the experimental results and the simulated results show that the segmentation results are very accurate. The Dice coefficients are improved by 70% to 80% compared with the Mean-Shift only method.

Keywords: gradient image, segmentation and extract, mean-shift algorithm, dictionary iearning

Procedia PDF Downloads 241

Authors: Hussaini Doko Ibrahim, Hamilton Cyprian Chinwenyi, Henrietta Nkem Ude

Abstract:

In this paper, efforts were made to examine and compare the algorithmic iterative solutions of the conjugate gradient method as against other methods such as Gauss-Seidel and Jacobi approaches for solving systems of linear equations of the form Ax=b, where A is a real n×n symmetric and positive definite matrix. We performed algorithmic iterative steps and obtained analytical solutions of a typical 3×3 symmetric and positive definite matrix using the three methods described in this paper (Gauss-Seidel, Jacobi, and conjugate gradient methods), respectively. From the results obtained, we discovered that the conjugate gradient method converges faster to exact solutions in fewer iterative steps than the two other methods, which took many iterations, much time, and kept tending to the exact solutions.

Keywords: conjugate gradient, linear equations, symmetric and positive definite matrix, gauss-seidel, Jacobi, algorithm

Procedia PDF Downloads 121

Authors: Selvakumar Ulaganathan, Ivo Couckuyt, Francesco Ferranti, Tom Dhaene, Eric Laermans

Abstract:

Variable-fidelity surrogate modelling offers an efficient way to approximate function data available in multiple degrees of accuracy each with varying computational cost. In this paper, a Kriging-based variable-fidelity surrogate modelling approach is introduced to approximate such deterministic data. Initially, individual Kriging surrogate models, which are enhanced with gradient data of different degrees of accuracy, are constructed. Then these Gradient enhanced Kriging surrogate models are strategically coupled using a recursive CoKriging formulation to provide an accurate surrogate model for the highest fidelity data. While, intuitively, gradient data is useful to enhance the accuracy of surrogate models, the primary motivation behind this work is to investigate if it is also worthwhile incorporating gradient data of varying degrees of accuracy.

Keywords: Kriging, CoKriging, Surrogate modelling, Variable- fidelity modelling, Gradients

Procedia PDF Downloads 530

Authors: Mahdieh Khalilinezhad, Silvana Dellepiane, Gianni Vernazza

Abstract:

The aim of the present work is to build a model based on tissue characterization that is able to discriminate pathological and non-pathological regions from three-phasic CT images. Based on feature selection in different phases, in this research, we design a neural network system that has optimal neuron number in a hidden layer. Our approach consists of three steps: feature selection, feature reduction, and classification. For each ROI, 6 distinct set of texture features are extracted such as first order histogram parameters, absolute gradient, run-length matrix, co-occurrence matrix, autoregressive model, and wavelet, for a total of 270 texture features. We show that with the injection of liquid and the analysis of more phases the high relevant features in each region changed. Our results show that for detecting HCC tumor phase3 is the best one in most of the features that we apply to the classification algorithm. The percentage of detection between these two classes according to our method, relates to first order histogram parameters with the accuracy of 85% in phase 1, 95% phase 2, and 95% in phase 3.

Keywords: multi-phasic liver images, texture analysis, neural network, hidden layer

Procedia PDF Downloads 242

Authors: A. Archibong-Eso, J. Shi, Y Baba, S. Alagbe, W. Yan, H. Yeung

Abstract:

This study presents over 100 experiments conducted in a 25.4 mm internal diameter (ID) horizontal pipeline. Oil viscosity ranging from 3.5 Pa.s–5.0 Pa.s are used with superficial velocities of oil and water ranging from 0.06 to 0.55 m/s and 0.01 m/s to 1.0 m/s, respectively. Pressure gradient measurements and flow pattern observations are discussed. Numerical simulation of some flow conditions is performed using the commercial CFD code ANSYS Fluent® and the simulation results are compared with experimental results. Results indicate that CFD numerical simulation performed moderately well in predicting the flow configurations observed in this study while discrepancies were observed in the pressure gradient predictions.

Keywords: flow patterns, plug, pressure gradient, rivulet

Procedia PDF Downloads 397

Authors: Morteza Mirhosseini, Amir B. Khoshnevis

Abstract:

The boundary layer separation and new active flow control of a NACA 0025 airfoil were studied experimentally. This new flow control is sometimes known as a co-flow jet (cfj) airfoil. This paper presents the fluctuating velocity in a wall jet over the co-flow jet airfoil subjected to an adverse pressure gradient and a curved surface. In these results, the fluctuating velocity at the inner part increasing by increased the angle of attack up to 12^o and this has due to the jet energized, while the angle of attack 20^o has different. The airfoil cord based Reynolds number has 10⁵.

Keywords: adverse pressure gradient, fluctuating velocity, wall jet, co-flow jet airfoil

Procedia PDF Downloads 461

Authors: Nishant Parashar, Sawan S. Sinha, Balaji Srinivasan

Abstract:

We perform an investigation of the unclosed terms in the evolution equation of the velocity gradient tensor (VGT) in compressible decaying turbulent flow. Velocity gradients in a compressible turbulent flow field influence several important nonlinear turbulent processes like cascading and intermittency. In an attempt to understand the dynamics of the velocity gradients various researchers have tried to model the unclosed terms in the evolution equation of the VGT. The existing models proposed for these unclosed terms have limited applicability. This is mainly attributable to the complex structure of the higher order gradient terms appearing in the evolution equation of VGT. We investigate these higher order gradients using the data from direct numerical simulation (DNS) of compressible decaying isotropic turbulent flow. The gas kinetic method aided with weighted essentially non-oscillatory scheme (WENO) based flow- reconstruction is employed to generate DNS data. By applying neural-network to the DNS data, we map the structure of the unclosed higher order gradient terms in the evolution of the equation of the VGT with VGT itself. We validate our findings by performing alignment based study of the unclosed higher order gradient terms obtained using the neural network with the strain rate eigenvectors.

Keywords: compressible turbulence, neural network, velocity gradient tensor, direct numerical simulation

Procedia PDF Downloads 138

Authors: H. J. Kim, G. C. Park, K. C. Kim, J. H. Shin

Abstract:

Cavities are frequently found beneath conduits on pile foundations in old embankments. Cavity reduces seepage length significantly and consequently causes piping failure of embankments. Case studies of embankment failures indicate that the relative settlement between ground and pile supported-concrete conduit was the main reason of the cavity. In this paper, an attempt to simulate the cavity-induced piping failure mechanism was made using finite element numerical method. Piping potential is examined by carrying out parametric study for influencing factors such as cavity length, water level, and flow conditions. The concentration of hydraulic gradient adjacent to cavity was found. It is found that the hydraulic gradient close to the cavity exceeds considerably the critical hydraulic gradient causing piping. Piping failure potential due to the existence of cavity is evaluated and contour map for the potential risk of an embankment for piping failure is proposed.

Keywords: cavity, hydraulic gradient, levee, piping

Procedia PDF Downloads 490

Authors: Romisaa Ali

Abstract:

The Policy Gradient approach is one of the deep reinforcement learning families that combines deep neural networks (DNN) with reinforcement learning RL to discover the optimum of the control problem through experience gained from the interaction between the robot and its surroundings. In contrast to earlier policy gradient algorithms, which were unable to handle these two types of error because of over-or under-estimation introduced by the deep neural network model, 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 PDF Downloads 141

Authors: Saman Ghaffarian, Ilgin Gökaşar

Abstract:

This paper presents a new automatic vehicle detection method from very high resolution aerial images to measure traffic density. The proposed method starts by extracting road regions from image using road vector data. Then, the road image is divided into equal sections considering resolution of the images. Gradient vectors of the road image are computed from edge map of the corresponding image. Gradient vectors on the each boundary of the sections are divided where the gradient vectors significantly change their directions. Finally, number of vehicles in each section is carried out by calculating the standard deviation of the gradient vectors in each group and accepting the group as vehicle that has standard deviation above predefined threshold value. The proposed method was tested in four very high resolution aerial images acquired from Istanbul, Turkey which illustrate roads and vehicles with diverse characteristics. The results show the reliability of the proposed method in detecting vehicles by producing 86% overall F1 accuracy value.

Keywords: aerial images, intelligent transportation systems, traffic density measurement, vehicle detection

Procedia PDF Downloads 358

Authors: Salim Ouchtati, Aissa Belmeguenai, Mouldi Bedda

Abstract:

In this study, we propose an offline system for the recognition of the fragmented handwritten numeric chains. Firstly, we realized a recognition system of the isolated handwritten digits, in this part; the study is based mainly on the evaluation of neural network performances, trained with the gradient backpropagation algorithm. The used parameters to form the input vector of the neural network are extracted from the binary images of the isolated handwritten digit by several methods: the distribution sequence, sondes application, the Barr features, and the centered moments of the different projections and profiles. Secondly, the study is extended for the reading of the fragmented handwritten numeric chains constituted of a variable number of digits. The vertical projection was used to segment the numeric chain at isolated digits and every digit (or segment) was presented separately to the entry of the system achieved in the first part (recognition system of the isolated handwritten digits).

Keywords: features extraction, handwritten numeric chains, image processing, neural networks

Procedia PDF Downloads 243

Authors: Liang Liu, Xiaopeng Luo

Abstract:

In this paper, we propose an accelerated stochastic gradient method with momentum. The momentum term is the weighted average of generated gradients, and the weights decay inverse proportionally with the iteration times. Stochastic gradient descent with momentum (SGDM) uses weights that decay exponentially with the iteration times to generate the momentum term. Using exponential decay weights, variants of SGDM with inexplicable and complicated formats have been proposed to achieve better performance. However, the momentum update rules of our method are as simple as that of SGDM. We provide theoretical convergence analyses, which show both the exponential decay weights and our inverse proportional decay weights can limit the variance of the parameter moving directly to a region. Experimental results show that our method works well with many practical problems and outperforms SGDM.

Keywords: exponential decay rate weight, gradient descent, inverse proportional decay rate weight, momentum

Procedia PDF Downloads 132

Authors: Pu Ying-Chih, Yang Yin-Ju, Hang Jian-Yi, Jang Guang-Way

Abstract:

Organic-Inorganic hybrid materials consisting of graded distributions of inorganic nano particles in organic polymer matrices were successfully prepared by the sol-gel process. Optical and surface properties of the resulting nano composites can be manipulated by changing their compositions and nano particle distribution gradients. Applications of gradient nano composite materials include sealants for LED packaging and screen lenses for smartphones. Optical transparency, prism coupler, TEM, SEM, Energy Dispersive X-ray Spectrometer (EDX), Izod impact strength, conductivity, pencil hardness, and thermogravimetric characterizations of the nano composites were performed and the results will be presented.

Keywords: Gradient, Hybrid, Nanocomposite, Organic-Inorganic

Procedia PDF Downloads 479

Authors: Souvik Phadikar, Nidul Sinha, Rajdeep Ghosh

Abstract:

In brainwaves research using electroencephalogram (EEG) signals, finding the most relevant and effective feature set for identification of activities in the human brain is a big challenge till today because of the random nature of the signals. The feature extraction method is a key issue to solve this problem. Finding those features that prove to give distinctive pictures for different activities and similar for the same activities is very difficult, especially for the number of activities. The performance of a classifier accuracy depends on this quality of feature set. Further, more number of features result in high computational complexity and less number of features compromise with the lower performance. In this paper, a novel idea of the selection of optimal feature set using a heuristically optimized deep autoencoder is presented. Using various feature extraction methods, a vast number of features are extracted from the EEG signals and fed to the autoencoder deep neural network. The autoencoder encodes the input features into a small set of codes. To avoid the gradient vanish problem and normalization of the dataset, a meta-heuristic search algorithm is used to minimize the mean square error (MSE) between encoder input and decoder output. To reduce the feature set into a smaller one, 4 hidden layers are considered in the autoencoder network; hence it is called Heuristically Optimized Deep Autoencoder (HO-DAE). In this method, no features are rejected; all the features are combined into the response of responses of the hidden layer. The results reveal that higher accuracy can be achieved using optimal reduced features. The proposed HO-DAE is also compared with the regular autoencoder to test the performance of both. The performance of the proposed method is validated and compared with the other two methods recently reported in the literature, which reveals that the proposed method is far better than the other two methods in terms of classification accuracy.

Keywords: autoencoder, brainwave signal analysis, electroencephalogram, feature extraction, feature selection, optimization

Procedia PDF Downloads 95

Authors: Tusar Kanti Dash, Ganapati Panda

Abstract:

The evaluation of speech quality and intelligence is critical to the overall effectiveness of the Speech Enhancement Algorithms. Several intrusive and non-intrusive measures are employed to calculate these parameters. Non-Intrusive Evaluation is most challenging as, very often, the reference clean speech data is not available. In this paper, a novel non-intrusive speech evaluation measure is proposed using audio features derived from the Stockwell transform. These features are used with the Light Gradient Boosting Machine for the effective prediction of speech quality and intelligibility. The proposed model is analyzed using noisy and reverberant speech from four databases, and the results are compared with the standard Intrusive Evaluation Measures. It is observed from the comparative analysis that the proposed model is performing better than the standard Non-Intrusive models.

Keywords: non-Intrusive speech evaluation, S-transform, light GBM, speech quality, and intelligibility

Procedia PDF Downloads 233

Authors: Jeong-Min Lee, Hyunwoo Song, Yonseok Kim, Junghan Yun, Jungin Byun, Jae-Mean Koo, Chang-Sung Seok

Abstract:

The Thermal Barrier Coating (TBC) prevents heat directly transferring from the high-temperature flame to the substrate. Top coat and bond coat compose the TBC and top coat consists of a ceramic and bond coat increases adhesion between the top coat and the substrate. The TBC technology drops the substrate surface temperature by about 150~200°C. In addition, the TBC system has a cooling system to lower the blade temperature by the air flow inside the blade. Then, as a result, the thermal gradient occurs inside the blade by cooling. Also, the internal stress occurs due to the difference in thermal expansion. In this paper, the finite element analyses (FEA) were performed and stress changes were derived according to the thermal gradient of the TBC system. The stress was increased due to the cooling, but difference of the stress between the top coat and bond coat was decreased. So, delamination in the interface between top coat and bond coat.

Keywords: gas turbine blade, Thermal Barrier Coating (TBC), thermal gradient, Finite Element Analysis (FEA)

Procedia PDF Downloads 582

Authors: Muhammad Danladi, Muhammad Bura Garba, Muhammad Yahaya, Dahiru Muhammad

Abstract:

Biomass is one of the primary sources of energy supply, which contributes to about 78% of Nigeria. In this work, a comparative analysis of the heating values of sawdust and rice husks in the thermal generation of electricity was carried out. In the study, different masses of biomass were used and the corresponding electromotive force in millivolts was obtained. A graph of e.m.f was plotted against the mass of each biomass and a gradient was obtained. Bar graphs were plotted to represent the values of e.m.f and masses of the biomass. Also, a graph of e.m.f against eating values of sawdust and rice husks was plotted, and in each case, as the e.m.f increases also, the heating values increases. The result shows that saw dust with 0.033Mv/g gradient and 3.5 points of intercept had the highest gradient, followed by rice husks with 0.026Mv/g gradient and 2.6 points of intercept. It is, therefore, concluded that sawdust is the most efficient of the two types of biomass in the thermal generation of electricity.

Keywords: biomass, electricity, thermal, generation

Procedia PDF Downloads 64

Authors: Armin Rahimi

Abstract:

In this study, we reveal co-seismic signals of two combined earthquakes, the 2006 Mw8.3 thrust and 2007 Mw8.1 normal fault earthquakes of the central Kuril Islands from GRACE observations. We compute monthly full gravitational gradient tensor in the local north-east-down frame for Kuril Islands earthquakes without spatial averaging and de-striping filters. Some of the gravitational gradient components (e.g. ΔVxx, ΔVxz) enhance high frequency components of the earth gravity field and reveal more details in spatial and temporal domain. Therefore that preseismic activity can be better illustrated. We show that the positive-negative-positive co-seismic ΔVxx due to the Kuril Islands earthquakes ranges from − 0.13 to + 0.11 milli Eötvös, and ΔVxz shows a positive-negative-positive pattern ranges from − 0.16 to + 0.13 milli Eötvös, agree well with seismic model predictions.

Keywords: GRACE observation, gravitational gradient changes, Kuril island earthquakes, PSGRN/PSCMP

Procedia PDF Downloads 242

Authors: Salim Ouchtati, Jean Sequeira, Mouldi Bedda

Abstract:

In this work we present an off line system for the recognition of the Arabic handwritten words of the Algerian departments. The study is based mainly on the evaluation of neural network performances, trained with the gradient back propagation algorithm. The used parameters to form the input vector of the neural network are extracted on the binary images of the handwritten word by several methods: the parameters of distribution, the moments centered of the different projections and the Barr features. It should be noted that these methods are applied on segments gotten after the division of the binary image of the word in six segments. The classification is achieved by a multi layers perceptron. Detailed experiments are carried and satisfactory recognition results are reported.

Keywords: handwritten word recognition, neural networks, image processing, pattern recognition, features extraction

Procedia PDF Downloads 485

Authors: Mohammed Belloufi, Rachid Benzine, Badreddine Sellami

Abstract:

Conjugate gradient methods is useful for solving large scale optimization problems in scientific and engineering computation, characterized by the simplicity of their iteration and their low memory requirements. It is well known that the search direction plays a main role in the line search method. In this paper, we propose a search direction with the Wolfe line search technique for solving unconstrained optimization problems. Under the above line searches and some assumptions, the global convergence properties of the given methods are discussed. Numerical results and comparisons with other CG methods are given.

Keywords: unconstrained optimization, conjugate gradient method, strong Wolfe line search, global convergence

Procedia PDF Downloads 390

Authors: Insaf Ajili, Malik Mallem, Jean-Yves Didier

Abstract:

Motion recognition from videos is actually a very complex task due to the high variability of motions. This paper describes the challenges of human motion recognition, especially motion representation step with relevant features. Our descriptor vector is inspired from Laban Movement Analysis method. We propose discriminative features using the Random Forest algorithm in order to remove redundant features and make learning algorithms operate faster and more effectively. We validate our method on MSRC-12 and UTKinect datasets.

Keywords: discriminative LMA features, features reduction, human motion recognition, random forest

Procedia PDF Downloads 163

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 PDF Downloads 111

Authors: Saba Ebrahimi, Saeed Ahmadian, Hedie Ashrafi

Abstract:

Selecting appropriate patients for surgery is one of the main issues in thoracic surgery (TS). Both short-term and long-term risks and benefits of surgery must be considered in the patient selection criteria. There are some limitations in the existing datasets of TS patients because of missing values of attributes and imbalanced distribution of survival classes. In this study, a novel ensemble architecture of deep learning networks is proposed based on stacking different linear and non-linear layers to deal with imbalance datasets. The categorical and numerical features are split using different layers with ability to shrink the unnecessary features. Then, after extracting the insight from the raw features, a novel biased-kernel layer is applied to reinforce the gradient of the minority class and cause the network to be trained better comparing the current methods. Finally, the performance and advantages of our proposed model over the existing models are examined for predicting patient survival after thoracic surgery using a real-life clinical data for lung cancer patients.

Keywords: deep learning, ensemble models, imbalanced classification, lung cancer, TS patient selection

Procedia PDF Downloads 114

Authors: Zhongjie Yu, Hancheng Yu

Abstract:

In this paper, we propose an effective method to detect concentric circles with imperfect edges. First, the gradient of edge pixel is coded and a 2-D lookup table is built to speed up normal generation. Then we take an accumulator to estimate the rough center and collect plausible edges of concentric circles through gradient and distance. Later, we take the contour-based method, which takes the contour and edge intersection, to pre-classify the edges. Finally, we use the extended RANSAC method to find all the candidate circles. The center of concentric circles is determined by the two circles with the highest concentricity. Experimental results demonstrate that the proposed method has both good performance and accuracy for the detection of concentric circles.

Keywords: concentric circle detection, gradient, contour, edge pre-classification, RANSAC

Procedia PDF Downloads 111

Authors: O. W. Lawal, L. O. Ahmed, Y. K. Ali

Abstract:

The unsteady MHD Poiseuille flow of a third grade fluid between two parallel horizontal nonconducting porous plates is studied with heat transfer. The two plates are fixed but maintained at different constant temperature with the Joule and viscous dissipation taken into consideration. The fluid motion is produced by a sudden uniform exponential decaying pressure gradient and external uniform magnetic field that is perpendicular to the plates. The momentum and energy equations governing the flow are solved numerically using Maple program. The effects of magnetic field and third grade fluid parameters on velocity and temperature profile are examined through several graphs.

Keywords: exponential decaying pressure gradient, MHD flow, Poiseuille flow, third grade fluid

Procedia PDF Downloads 449

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 PDF Downloads 398

Authors: Bastien Batardière, Joon Kwon

Abstract:

For finite-sum optimization, variance-reduced gradient methods (VR) compute at each iteration the gradient of a single function (or of a mini-batch), and yet achieve faster convergence than SGD thanks to a carefully crafted lower-variance stochastic gradient estimator that reuses past gradients. Another important line of research of the past decade in continuous optimization is the adaptive algorithms such as AdaGrad, that dynamically adjust the (possibly coordinate-wise) learning rate to past gradients and thereby adapt to the geometry of the objective function. Variants such as RMSprop and Adam demonstrate outstanding practical performance that have contributed to the success of deep learning. In this work, we present AdaLVR, which combines the AdaGrad algorithm with loopless variance-reduced gradient estimators such as SAGA or L-SVRG that benefits from a straightforward construction and a streamlined analysis. We assess that AdaLVR inherits both good convergence properties from VR methods and the adaptive nature of AdaGrad: in the case of L-smooth convex functions we establish a gradient complexity of O(n + (L + √ nL)/ε) without prior knowledge of L. Numerical experiments demonstrate the superiority of AdaLVR over state-of-the-art methods. Moreover, we empirically show that the RMSprop and Adam algorithm combined with variance-reduced gradients estimators achieve even faster convergence.

Keywords: convex optimization, variance reduction, adaptive algorithms, loopless

Procedia PDF Downloads 31

Authors: Ian Marc G. Cabugsa, Kim Ryan A. Won

Abstract:

Fast, reliable and simultaneous HPLC analysis of theobromine and caffeine in cacao and cocoa products was optimized in this study. The samples tested were raw, fermented, and roasted cacao beans as well as commercially available cocoa products. The HPLC analysis was carried out using step gradient solvent system with acetonitrile and water buffered with H3PO4 as the mobile phase. The HPLC system was optimized using 273 nm wavelength at 35 °C for the column temperature with a flow rate of 1.0 mL/min. Using this method, the theobromine percent recovery mean, Limit of Detection (LOD) and Limit of Quantification (LOQ) is 118.68(±3.38)%, 0.727 and 1.05 respectively. The percent recovery mean, LOD and LOQ for caffeine is 105.53(±3.25)%, 2.42 and 3.50 respectively. The inter-day and intra-day precision for theobromine is 4.31% and 4.48% respectively, while 7.02% and 7.03% was for caffeine respectively. Compared to the standard method in AOAC using methanol in isocratic solvent system, the results of the study produced lesser chromatogram noise with emphasis on theobromine and caffeine. The method is readily usable for cacao and cocoa substances analyses using HPLC with step gradient capability.

Keywords: cacao, caffeine, HPLC, step gradient solvent system, theobromine

Procedia PDF Downloads 258

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: steepest descent, line search, iteration, running time, unconstrained optimization, convergence

Procedia PDF Downloads 520