Support vector machine (SVM) is a popular supervised learning algorithm based on margin maximization. It has a high training cost and does not scale well to a large number of data points. We propose a multiresolution algorithm MRH-SVM that trains SVM on a hierarchical data aggregation structure, which also serves as a common data input to other learning algorithms. The proposed algorithm learns SVM models using high-level data aggregates and only visits data aggregates at more detailed levels where support vectors reside. In addition to performance improvements, the algorithm has advantages such as the ability to handle data streams and datasets with imbalanced classes. Experimental results show significant performance improvements in compa

In the present research a new test rig has been proposed to be suitable for different cyclic loads such as cyclic bending, cyclic torsion, proportional and non proportional loads. In this work the efforts were concentrated on the cyclic bending loads concerning cracked pipes with or without internal pulsing pressure to study crack propagation in small bore pipes (up to 1'') for transverse or inclined cracks. The rig simulates the real service conditions under different stresses by means the least dangerous case will be suggested, so the experiments were considered for copper pipe, and the results have been tabulated and drawn to demonstrate the crack growth behavior as well as to justify the outcomes practically, consequently the durabil

Due to the great evolution in digital commercial cameras, several studies have addressed the using of such cameras in different civil and close-range applications such as 3D models generation. However, previous studies have not discussed a precise relationship between a camera resolution and the accuracy of the models generated based on images of this camera. Therefore the current study aims to evaluate the accuracy of the derived 3D buildings models captured by different resolution cameras. The digital photogrammetric methods were devoted to derive 3D models using the data of various resolution cameras and analyze their accuracies. This investigation involves selecting three different resolution cameras (low, medium and

This paper deals with the nonlinear large-angle bending dynamic analysis of curved beams which investigated by modeling wave’s transmission along curved members. The approach depends on the wave propagation in one-dimensional structural element using the method of characteristics. The method of characteristics (MOC) is found to be a suitable method for idealizing the wave propagation inside structural systems. Timoshenko’s beam theory, which includes transverse shear deformation and rotary inertia effects, is adopted in the analysis. Only geometrical non-linearity is considered in this study and the material is assumed to be linearly elastic. Different boundary conditions and loading cases are examined.

From the results obtai

Abstract

Theoretical and experimental methodologies were assessed to test curved beam made of layered   composite material. The maximum stress and maximum deflection were computed for each layer and the effect of radius of curvature and curve shape on them. Because of the increase of the use of composite materials in aircraft structures and the renewed interest in these types of problems, the presented theoretical assessment was made using three different approaches: curved beam theory and an approximate 2D strength of material equations and finite element method (FEM) analysis by ANSYS 14.5 program for twelve cases of multi-layered cylindrical shell panel differs in fibe

Many academics have concentrated on applying machine learning to retrieve information from databases to enable researchers to perform better. A difficult issue in prediction models is the selection of practical strategies that yield satisfactory forecast accuracy. Traditional software testing techniques have been extended to testing machine learning systems; however, they are insufficient for the latter because of the diversity of problems that machine learning systems create. Hence, the proposed methodologies were used to predict flight prices. A variety of artificial intelligence algorithms are used to attain the required, such as Bayesian modeling techniques such as Stochastic Gradient Descent (SGD), Adaptive boosting (ADA), Decision Tre

Many academics have concentrated on applying machine learning to retrieve information from databases to enable researchers to perform better. A difficult issue in prediction models is the selection of practical strategies that yield satisfactory forecast accuracy. Traditional software testing techniques have been extended to testing machine learning systems; however, they are insufficient for the latter because of the diversity of problems that machine learning systems create. Hence, the proposed methodologies were used to predict flight prices. A variety of artificial intelligence algorithms are used to attain the required, such as Bayesian modeling techniques such as Stochastic Gradient Descent (SGD), Adaptive boosting (ADA), Deci

Botnet detection develops a challenging problem in numerous fields such as order, cybersecurity, law, finance, healthcare, and so on. The botnet signifies the group of co-operated Internet connected devices controlled by cyber criminals for starting co-ordinated attacks and applying various malicious events. While the botnet is seamlessly dynamic with developing counter-measures projected by both network and host-based detection techniques, the convention techniques are failed to attain sufficient safety to botnet threats. Thus, machine learning approaches are established for detecting and classifying botnets for cybersecurity. This article presents a novel dragonfly algorithm with multi-class support vector machines enabled botnet