scholarly journals A novel pooling layer based on gaussian function with wavelet transform

Author(s):  
Aqeel M. Hamad alhussainy ◽  
Ammar D. Jasim

<p>Convolution represent basic layer in the convolutional neural network, but it<br />can result in big size of the data, which may increase the complexity of the<br />network. Different pooling methods are used to perform down sample these<br />data. In this paper, we have proposed a novel pooling method by using<br />Gaussian function to determine the wavelet filter coefficients . At first, the<br />basic statistics are determined for each pool size of the signal, then Gaussian<br />probability distribution function is determined . According to the procedure<br />of extracting the features , three methods are proposed , the first method is<br />used the normalized values of basic statistics as wavelet filter to be<br />multiplied by original signal, the second method used the determined<br />statistics as features of the original signal ,then multiplied it with constant<br />wavelet filter based on Gaussian ,while the third method is similar to first<br />method, except it depend on entire signal instead of each pool size. The<br />proposed methods are combined with other standard methods such as max<br />and pooling. The experiments are performed on different datasets , and the<br />results show that the proposed methods perform or outperform other<br />methods and can increase performance of the (CNN).</p>

2006 ◽  
Vol 306-308 ◽  
pp. 1491-1496
Author(s):  
Chun He Yang ◽  
Guibin Wang ◽  
Liming Yin

Based on the field geology survey data, the probability distribution function of joint aperture was analyzed and the availability of the fitting data was verified. The joint aperture data observed the previous fitted function can be generated in the process of joint simulation. The seepage of rock mass is proportional to the third power of joint’s aperture. It is much more reasonable and feasible to analyze the permeability tensor of rock mass with the previous generated data. At the same time, the method comprehensively considered the effect of joint’s orientation, aperture, scale and their distribution.


2017 ◽  
Vol 74 (9) ◽  
pp. 2921-2935 ◽  
Author(s):  
Mikhail D. Alexandrov ◽  
Alexander Marshak

Abstract In the third part of the “Cellular Statistical Models of Broken Cloud Fields” series the cloud statistics formalism developed in the first two parts is interpreted in terms of the theory of Markov processes. The master matrix introduced in this study is a unifying generalization of both the cloud fraction probability distribution function (PDF) and the Markovian transition probability matrix. To illustrate the new concept, the master matrix is used for computation of the moments of the cloud fraction PDF—in particular, the variance—which until now has not been analytically derived in the framework of the authors’ previous work. This paper also serves as a bridge to the proposed future studies of the effects of sampling and averaging on satellite-based cloud masks.


Electronics ◽  
2021 ◽  
Vol 10 (11) ◽  
pp. 1248
Author(s):  
Rafia Nishat Toma ◽  
Cheol-Hong Kim ◽  
Jong-Myon Kim

Condition monitoring is used to track the unavoidable phases of rolling element bearings in an induction motor (IM) to ensure reliable operation in domestic and industrial machinery. The convolutional neural network (CNN) has been used as an effective tool to recognize and classify multiple rolling bearing faults in recent times. Due to the nonlinear and nonstationary nature of vibration signals, it is quite difficult to achieve high classification accuracy when directly using the original signal as the input of a convolution neural network. To evaluate the fault characteristics, ensemble empirical mode decomposition (EEMD) is implemented to decompose the signal into multiple intrinsic mode functions (IMFs) in this work. Then, based on the kurtosis value, insignificant IMFs are filtered out and the original signal is reconstructed with the rest of the IMFs so that the reconstructed signal contains the fault characteristics. After that, the 1-D reconstructed vibration signal is converted into a 2-D image using a continuous wavelet transform with information from the damage frequency band. This also transfers the signal into a time-frequency domain and reduces the nonstationary effects of the vibration signal. Finally, the generated images of various fault conditions, which possess a discriminative pattern relative to the types of faults, are used to train an appropriate CNN model. Additionally, with the reconstructed signal, two different methods are used to create an image to compare with our proposed image creation approach. The vibration signal is collected from a self-designed testbed containing multiple bearings of different fault conditions. Two other conventional CNN architectures are compared with our proposed model. Based on the results obtained, it can be concluded that the image generated with fault signatures not only accurately classifies multiple faults with CNN but can also be considered as a reliable and stable method for the diagnosis of fault bearings.


2021 ◽  
Vol 5 (1) ◽  
pp. 46
Author(s):  
Mostafa Abotaleb ◽  
Tatiana Makarovskikh

COVID-19 is one of the biggest challenges that countries face at the present time, as infections and deaths change daily and because this pandemic has a dynamic spread. Our paper considers two tasks. The first one is to develop a system for modeling COVID-19 based on time-series models due to their accuracy in forecasting COVID-19 cases. We developed an “Epidemic. TA” system using R programming for modeling and forecasting COVID-19 cases. This system contains linear (ARIMA and Holt’s model) and non-linear (BATS, TBATS, and SIR) time-series models and neural network auto-regressive models (NNAR), which allows us to obtain the most accurate forecasts of infections, deaths, and vaccination cases. The second task is the implementation of our system to forecast the risk of the third wave of infections in the Russian Federation.


Soft Matter ◽  
2020 ◽  
Author(s):  
Ulices Que-Salinas ◽  
Pedro Ezequiel Ramirez-Gonzalez ◽  
Alexis Torres-Carbajal

In this work we implement a machine learning method to predict the thermodynamic state of a liquid using only its microscopic structure provided by the radial distribution function (RDF). The...


Radiocarbon ◽  
2016 ◽  
Vol 59 (5) ◽  
pp. 1623-1627 ◽  
Author(s):  
Ron W Reimer ◽  
Paula J Reimer

AbstractA regional offset (ΔR) from the marine radiocarbon calibration curve is widely used in calibration software (e.g. CALIB, OxCal) but often is not calculated correctly. While relatively straightforward for known-age samples, such as mollusks from museum collections or annually banded corals, it is more difficult to calculate ΔR and the uncertainty in ΔR for 14C dates on paired marine and terrestrial samples. Previous researchers have often utilized classical intercept methods that do not account for the full calibrated probability distribution function (pdf). Recently, Soulet (2015) provided R code for calculating reservoir ages using the pdfs, but did not address ΔR and the uncertainty in ΔR. We have developed an online application for performing these calculations for known-age, paired marine and terrestrial 14C dates and U-Th dated corals. This article briefly discusses methods that have been used for calculating ΔR and the uncertainty and describes the online program deltar, which is available free of charge.


2021 ◽  
Vol 14 (16) ◽  
Author(s):  
Adnan A. Ismael ◽  
Saleh J. Suleiman ◽  
Raid Rafi Omar Al-Nima ◽  
Nadhir Al-Ansari

AbstractCylindrical weir shapes offer a steady-state overflow pattern, where the type of weirs can offer a simple design and provide the ease-to-pass floating debris. This study considers a coefficient of discharge (Cd) prediction for oblique cylindrical weir using three diameters, the first is of D1 = 0.11 m, the second is of D2 = 0.09 m, and the third is of D3 = 0.06.5 m, and three inclination angles with respect to channel axis, the first is of θ1 = 90 ͦ, the second is of θ2 = 45 ͦ, and the third is of θ3 = 30 ͦ. The Cd values for total of 56 experiments are estimated by using the radial basis function network (RBFN), in addition of comparing that with the back-propagation neural network (BPNN) and cascade-forward neural network (CFNN). Root mean square error (RMSE), mean square error (MSE), and correlation coefficient (CC) statics are used as metrics measurements. The RBFN attained superior performance comparing to the other neural networks of BPNN and CFNN. It is found that, for the training stage, the RBFN network benchmarked very small RMSE and MSE values of 1.35E-12 and 1.83E-24, respectively and for the testing stage, it also could benchmark very small RMSE and MSE values of 0.0082 and 6.80E-05, respectively.


2012 ◽  
Vol 706 ◽  
pp. 118-149 ◽  
Author(s):  
Dennis P. M. van Gils ◽  
Sander G. Huisman ◽  
Siegfried Grossmann ◽  
Chao Sun ◽  
Detlef Lohse

AbstractStrongly turbulent Taylor–Couette flow with independently rotating inner and outer cylinders with a radius ratio of $\eta = 0. 716$ is experimentally studied. From global torque measurements, we analyse the dimensionless angular velocity flux ${\mathit{Nu}}_{\omega } (\mathit{Ta}, a)$ as a function of the Taylor number $\mathit{Ta}$ and the angular velocity ratio $a= \ensuremath{-} {\omega }_{o} / {\omega }_{i} $ in the large-Taylor-number regime $1{0}^{11} \lesssim \mathit{Ta}\lesssim 1{0}^{13} $ and well off the inviscid stability borders (Rayleigh lines) $a= \ensuremath{-} {\eta }^{2} $ for co-rotation and $a= \infty $ for counter-rotation. We analyse the data with the common power-law ansatz for the dimensionless angular velocity transport flux ${\mathit{Nu}}_{\omega } (\mathit{Ta}, a)= f(a)\hspace{0.167em} {\mathit{Ta}}^{\gamma } $, with an amplitude $f(a)$ and an exponent $\gamma $. The data are consistent with one effective exponent $\gamma = 0. 39\pm 0. 03$ for all $a$, but we discuss a possible $a$ dependence in the co- and weakly counter-rotating regimes. The amplitude of the angular velocity flux $f(a)\equiv {\mathit{Nu}}_{\omega } (\mathit{Ta}, a)/ {\mathit{Ta}}^{0. 39} $ is measured to be maximal at slight counter-rotation, namely at an angular velocity ratio of ${a}_{\mathit{opt}} = 0. 33\pm 0. 04$, i.e. along the line ${\omega }_{o} = \ensuremath{-} 0. 33{\omega }_{i} $. This value is theoretically interpreted as the result of a competition between the destabilizing inner cylinder rotation and the stabilizing but shear-enhancing outer cylinder counter-rotation. With the help of laser Doppler anemometry, we provide angular velocity profiles and in particular identify the radial position ${r}_{n} $ of the neutral line, defined by $ \mathop{ \langle \omega ({r}_{n} )\rangle } \nolimits _{t} = 0$ for fixed height $z$. For these large $\mathit{Ta}$ values, the ratio $a\approx 0. 40$, which is close to ${a}_{\mathit{opt}} = 0. 33$, is distinguished by a zero angular velocity gradient $\partial \omega / \partial r= 0$ in the bulk. While for moderate counter-rotation $\ensuremath{-} 0. 40{\omega }_{i} \lesssim {\omega }_{o} \lt 0$, the neutral line still remains close to the outer cylinder and the probability distribution function of the bulk angular velocity is observed to be monomodal. For stronger counter-rotation the neutral line is pushed inwards towards the inner cylinder; in this regime the probability distribution function of the bulk angular velocity becomes bimodal, reflecting intermittent bursts of turbulent structures beyond the neutral line into the outer flow domain, which otherwise is stabilized by the counter-rotating outer cylinder. Finally, a hypothesis is offered allowing a unifying view and consistent interpretation for all these various results.


Sign in / Sign up

Export Citation Format

Share Document