Current Issue : October - December Volume : 2017 Issue Number : 4 Articles : 5 Articles
Many models of neural networks have been extended to complex-valued neural networks. A complex-valued Hopfield neural\nnetwork (CHNN) is a complex-valued version of a Hopfield neural network. Complex-valued neurons can represent multistates,\nand CHNNs are available for the storage of multilevel data, such as gray-scale images. The CHNNs are often trapped into the\nlocal minima, and their noise tolerance is low. Lee improved the noise tolerance of the CHNNs by detecting and exiting the local\nminima. In the present work, we propose a new recall algorithm that eliminates the local minima.We show that our proposed recall\nalgorithm not only accelerated the recall but also improved the noise tolerance through computer simulations....
Extreme Learning Machine (ELM) is a fast-learning algorithm for a single-hidden layer feedforward neural network (SLFN). It\noften has good generalization performance. However, there are chances that it might overfit the training data due to having more\nhidden nodes than needed. To address the generalization performance, we use a heterogeneous ensemble approach. We propose\nan Advanced ELM Ensemble (AELME) for classification, which includes Regularized-ELM, ...
Digit Recognition is an essential element of the process of scanning and converting\ndocuments into electronic format. In this work, a new Multiple-Cell\nSize (MCS) approach is being proposed for utilizing Histogram of Oriented\nGradient (HOG) features and a Support Vector Machine (SVM) based classifier\nfor efficient classification of Handwritten Digits. The HOG based technique\nis sensitive to the cell size selection used in the relevant feature extraction\ncomputations. Hence a new MCS approach has been used to perform\nHOG analysis and compute the HOG features. The system has been tested on\nthe Benchmark MNIST Digit Database of handwritten digits and a classification\naccuracy of 99.36% has been achieved using an Independent Test set\nstrategy. A Cross-Validation analysis of the classification system has also been\nperformed using the 10-Fold Cross-Validation strategy and a 10-Fold classification\naccuracy of 99.26% has been obtained. The classification performance\nof the proposed system is superior to existing techniques using complex procedures\nsince it has achieved at par or better results using simple operations in\nboth the Feature Space and in the Classifier Space. The plots of the system�s\nConfusion Matrix and the Receiver Operating Characteristics (ROC) show\nevidence of the superior performance of the proposed new MCS HOG and\nSVM based digit classification system....
In the data mining, the analysis of high-dimensional data is a critical but thorny research topic.TheLASSO (least absolute shrinkage\nand selection operator) algorithm avoids the limitations, which generally employ stepwise regression with information criteria to\nchoose the optimal model, existing in traditional methods. The improved-LARS (Least Angle Regression) algorithm solves the\nLASSO effectively. This paper presents an improved-LARS algorithm, which is constructed on the basis ofmultidimensional weight\nand intends to solve the problems in LASSO. Specifically, in order to distinguish the impact of each variable in the regression,\nwe have separately introduced part of principal component analysis (Part_PCA), Independent Weight evaluation, and CRITIC,\ninto our proposal.We have explored that these methods supported by our proposal change the regression track by weighted every\nindividual, to optimize the approach direction, aswell as the approach variable selection. As a consequence, our proposed algorithm\ncan yield better results in the promise direction. Furthermore, we have illustrated the excellent property of LARS algorithm based\non multidimensional weight by the Pima Indians Diabetes. The experiment results show an attractive performance improvement\nresulting from the proposed method, compared with the improved-LARS, when they are subjected to the same threshold value....
This study investigates an adaptive-weighted instanced-based learning, for the prediction of the ultimate punching shear capacity\n(UPSC) of fiber-reinforced polymer- (FRP-) reinforced slabs.Theconcept of the newmethod is to employ theDifferential Evolution\nto construct an adaptive instance-based regressionmodel.The performance of the proposedmodel is compared to thoseofArtificial\nNeural Network (ANN) and traditional formula-based methods. A dataset which contains the testing results of FRP-reinforced\nconcrete slabs has been collected to establish and verify new approach. This study shows that the investigated instance-based\nregression model is capable of delivering the prediction result which is far more accurate than traditional formulas and very\ncompetitivewith the black-box approach ofANN. Furthermore, the proposed adaptive-weighted instanced-based learning provides\na means for quantifying the relevancy of each factor used for the prediction of UPSC of FRP-reinforced slabs....
Loading....