A three-stage learning algorithm for deep multilayer perceptron (DMLP) with effective weight initialisation based on sparse auto-encoder is proposed in this paper, which aims to overcome difficulties in training deep neural networks with limited training data in high-dimensional feature space. At the first stage, unsupervised learning is adopted using sparse auto-encoder to obtain the initial weights of the feature extraction layers of the DMLP. At the second stage, error back-propagation is used to train the DMLP by fixing the weights obtained at the first stage for its feature extraction layers. At the third stage, all the weights of the DMLP obtained at the second stage are refined by error back-propagation. Network structures an

This paper generalizes and improves the results of Margenstren, by proving that the number of -practical numbers  which is defined by   has a lower bound in terms of . This bound is more sharper than Mangenstern bound when  Further general results are given for the existence of -practical numbers, by proving that the interval contains a -practical for all

Various speech enhancement Algorithms (SEA) have been developed in the last few decades. Each algorithm has its advantages and disadvantages because the speech signal is affected by environmental situations. Distortion of speech results in the loss of important features that make this signal challenging to understand. SEA aims to improve the intelligibility and quality of speech that different types of noise have degraded. In most applications, quality improvement is highly desirable as it can reduce listener fatigue, especially when the listener is exposed to high noise levels for extended periods (e.g., manufacturing). SEA reduces or suppresses the background noise to some degree, sometimes called noise suppression alg

This work presents a symmetric cryptography coupled with Chaotic NN , the encryption algorithm process the data as a blocks and it consists of multilevel( coding of character, generates array of keys (weights),coding of text and chaotic NN ) , also the decryption process consists of multilevel (generates array of keys (weights),chaotic NN, decoding of text and decoding of character).Chaotic neural network is used as a part of the proposed system with modifying on it ,the keys that are used in chaotic sequence are formed by proposed key generation algorithm .The proposed algorithm appears efficiency during the execution time where it can encryption and decryption long messages by short time and small memory (chaotic NN offer capacity of m

This paper proposes improving the structure of the neural controller based on the identification model for nonlinear systems. The goal of this work is to employ the structure of the Modified Elman Neural Network (MENN) model into the NARMA-L2 structure instead of Multi-Layer Perceptron (MLP) model in order to construct a new hybrid neural structure that can be used as an identifier model and a nonlinear controller for the SISO linear or nonlinear systems. Two learning algorithms are used to adjust the parameters weight of the hybrid neural structure with its serial-parallel configuration; the first one is supervised learning algorithm based Back Propagation Algorithm (BPA) and the second one is an intelligent algorithm n

Image compression plays an important role in reducing the size and storage of data while increasing the speed of its transmission through the Internet significantly. Image compression is an important research topic for several decades and recently, with the great successes achieved by deep learning in many areas of image processing, especially image compression, and its use is increasing Gradually in the field of image compression. The deep learning neural network has also achieved great success in the field of processing and compressing various images of different sizes. In this paper, we present a structure for image compression based on the use of a Convolutional AutoEncoder (CAE) for deep learning, inspired by the diversity of human eye

 Bored piles settlement behavior under vertical loaded is the main factor that affects the design requirements of single or group of piles in soft soils. The estimation of bored pile settlement is a complicated problem because it depends upon many factors which may include ground conditions, validation of bored pile design method through testing and validation of theoretical or numerical prediction of the settlement value. In this study, a prototype single and bored pile group model of arrangement (1*1, 1*2 and 2*2) for total length to diameter ratios (L/D) is 13.33 and clear spacing three times of diameter, subjected to vertical axial loads. The bored piles model used for the test was 2000

Non uniform channelization is a crucial task in cognitive radio receivers for obtaining separate channels from the digitized wideband input signal at different intervals of time. The two main requirements in the channelizer are reconfigurability and low complexity. In this paper, a reconfigurable architecture based on a combination of Improved Coefficient Decimation Method (ICDM) and Coefficient Interpolation Method (CIM) is proposed. The proposed Hybrid Coefficient Decimation-Interpolation Method (HCDIM) based filter bank (FB) is able to realize the same number of channels realized using (ICDM) but with a maximum decimation factor divided by the interpolation factor (L), which leads to less deterioration in stop band at