Application of Natural Neighbor-based Algorithm on Oversampling SMOTE Algorithms

Author(s):  
Chutimet Srinilta ◽  
Sivakorn Kanharattanachai
Keyword(s):  
Author(s):  
N. A. Nascimento ◽  
J. Belinha ◽  
R. M. Natal Jorge ◽  
D. E. S. Rodrigues

Cellular solid materials are progressively becoming more predominant in lightweight structural applications as more technologies realize these materials can be improved in terms of performance, quality control, repeatability and production costs, when allied with fast developing manufacturing technologies such as Additive Manufacturing. In parallel, the rapid advances in computational power and the use of new numerical methods, such as Meshless Methods, in addition to the Finite Element Method (FEM) are highly beneficial and allow for more accurate studies of a wide range of topologies associated with the architecture of cellular solid materials. Since these materials are commonly used as the cores of sandwich panels, in this work, two different topologies were designed — conventional honeycombs and re-entrant honeycombs — for 7 different values of relative density, and tested on the linear-elastic domain, in both in-plane directions, using the Natural Neighbor Radial Point Interpolation Method (NNRPIM), a newly developed meshless method, and the Finite Element Method (FEM) for comparison purposes.


2019 ◽  
Vol 11 (21) ◽  
pp. 116-126
Author(s):  
Israa Jameel Muhsin

DEMs, thus, simply regular grids of elevation measurements over the land surface.The aim of the present work is to produce high resolution DEM for certain investigated region (i.e. Baghdad University Campus\ college of science). The easting and northing of 90 locations, including the ground-base and buildings of the studied area, have been obtained by field survey using global positioning system (GPS). The image of the investigated area has been extracted from Quick-Bird satellite sensor (with spatial resolution of 0.6 m). It has been geo-referenced and rectified  using 1st order polynomial transformation. many interpolation methods have been used to estimate the elevation such as ordinary Kriging, inverse distance weighted (IDW) and  natural neighbor methods. The mosaic  algorithm has then been applied between the base and building layers of studied area in order to perform the final DEM. The accuracy assessments of the interpolation methods have been calculated using the root-mean-square-error (RMSE) criterion. Finally, the estimated DEMs have been used to constructing 3-D views of the original image.


Author(s):  
Lijun Yang ◽  
Qingsheng Zhu ◽  
Jinlong Huang ◽  
Dongdong Cheng ◽  
Cheng Zhang

Instance reduction is aimed at reducing prohibitive computational costs and the storage space for instance-based learning. The most frequently used methods include the condensation and edition approaches. Condensation method removes the patterns far from the decision boundary and do not contribute to better classification accuracy, while edition method removes noisy patterns to improve the classification accuracy. In this paper, a new hybrid algorithm called instance reduction algorithm based on natural neighbor and nearest enemy is presented. At first, an edition algorithm is proposed to filter noisy patterns and smooth the class boundaries by using natural neighbor. The main advantage of the algorithm is that it does not require any user-defined parameters. Then, using a new condensation method based on nearest enemy to reduce instances far from decision line. Through this algorithm, interior instances are discarded. Experiments show that the hybrid approach effectively reduces the number of instances while achieves higher classification accuracy along with competitive algorithms.


Sign in / Sign up

Export Citation Format

Share Document