Image classification is the process of finding common features in images from various classes and applying them to categorize and label them. The main problem of the image classification process is the abundance of images, the high complexity of the data, and the shortage of labeled data, presenting the key obstacles in image classification. The cornerstone of image classification is evaluating the convolutional features retrieved from deep learning models and training them with machine learning classifiers. This study proposes a new approach of “hybrid learning” by combining deep learning with machine learning for image classification based on convolutional feature extraction using the VGG-16 deep learning model and seven class

An analytical model in the form of a hyperbolic function has been suggested for the axial potential distribution of an electrostatic einzel lens. With the aid of this hyperbolic model the relative optical parameters have been computed and investigated in detail as a function of the electrodes voltage ratio for various trajectories of an accelerated charged-particles beam. The electrodes voltage ratio covered a wide range where the lens may be operated at accelerating and decelerating modes. The results have shown that the proposed hyperbolic field has the advantages of producing low aberrations under various magnification conditions and operational modes. The electrodes profile and their three-dimensional diagram have been determined whi

Image classification is the process of finding common features in images from various classes and applying them to categorize and label them. The main problem of the image classification process is the abundance of images, the high complexity of the data, and the shortage of labeled data, presenting the key obstacles in image classification. The cornerstone of image classification is evaluating the convolutional features retrieved from deep learning models and training them with machine learning classifiers. This study proposes a new approach of “hybrid learning” by combining deep learning with machine learning for image classification based on convolutional feature extraction using the VGG-16 deep learning model and seven class

In this paper, a new equivalent lumped parameter model is proposed for describing the vibration of beams under the moving load effect. Also, an analytical formula for calculating such vibration for low-speed loads is presented. Furthermore, a MATLAB/Simulink model is introduced to give a simple and accurate solution that can be used to design beams subjected to any moving loads, i.e., loads of any magnitude and speed. In general, the proposed Simulink model can be used much easier than the alternative FEM software, which is usually used in designing such beams. The obtained results from the analytical formula and the proposed Simulink model were compared with those obtained from Ansys R19.0, and very good agreement has been shown. I

The challenge to incorporate usability evaluation values and practices into agile development process is not only persisting but also systemic. Notable contributions of researchers have attempted to isolate and close the gaps between both fields, with the aim of developing usable software. Due to the current absence of a reference model that specifies where and how usability activities need to be considered in the agile development process. This paper proposes a model for identifying appropriate usability evaluation methods alongside the agile development process. By using this model, the development team can apply usability evaluations at the right time at the right place to get the necessary feedback from the end-user. Verificatio

The problem of the high peak to average ratio (PAPR) in OFDM signals is investigated with a brief presentation of the various methods used to reduce the PAPR with special attention to the clipping method. An alternative approach of clipping is presented, where the clipping is performed right after the IFFT stage unlike the conventional clipping that is performed in the power amplifier stage, which causes undesirable out of signal band spectral growth. In the proposed method, there is clipping of samples not clipping of wave, therefore, the spectral distortion is avoided. Coding is required to correct the errors introduced by the clipping and the overall system is tested for two types of modulations, the QPSK as a constant amplitude modul

The thermal method was used to produce silicoaluminophosphate (SAPO-11) with different amounts of carbon nanotubes (CNT). XRD, nitrogen adsorption-desorption, SEM, AFM, and FTIR were used to characterize the prepared catalyst. It was discovered that adding CNT increased the crystallinity of the synthesize SAPO-11 at all the temperatures which studied, wile the maximum surface area was 179.54 m2/g obtained at 190°C with 7.5 percent of CNT with a pore volume of 0.317 cm3/g ,and with nano-particles with average particle diameter of 24.8 nm, while the final molar composition of the prepared SAPO-11 was (Al2O3:0.93P2O5:0.414SiO2).

The biosorption of Pb (II), Cd (II), and Hg (II) from simulated aqueous solutions using baker’s yeast biomass was investigated. Batch type experiments were carried out to find the equilibrium isotherm data for each component (single, binary, and ternary), and the adsorption rate constants. Kinetics pseudo-first and second order rate models applied to the adsorption data to estimate the rate constant for each solute, the results showed that the Cd (II), Pb (II), and Hg (II) uptake process followed the pseudo-second order rate model with (R2) 0.963, 0.979, and 0.960 respectively. The equilibrium isotherm data were fitted with five theoretical models. Langmuir model provides the best fitting for the experimental results with (R2) 0.992, 0