Hardware-friendly Higher-Order Neural Network Training using Distributed Evolutionary Algorithms

Artificial neural networks (ANN) have been widely used in the field of data classification. Normally, training of neural network is applied with the traditional back propagation technique. As, this approach has various drawbacks, training of neural network is done with Particle Swarm Optimization (PSO). PSO has been widely used to solve the diverse kind of optimization problems. Population initialization performs a significant role in meta-heuristic algorithms. This paper describes a new initialization population approach Log Logistic termed as PSOLL-NN to create the initialization of the swarm. The proposed algorithm has been tested for weight optimization of feed forward neural network; and compared with back propagation Algorithm (BPA), standard PSO (PSO-NN), PSO initialized with Halton Sequence (PSOH-NN), Torus sequence (PSOT-NN) and Sobol sequence (PSOS-NN). The experimental results show that the proposed technique performed exceptionally better than the other traditional techniques. Moreover, the outcome of our work presents a foresight that how the proposed initialization technique can be used as an efficient alternative to standard training approaches for the data classification problems.

On the training and performance of high-order neural networks

ABSTRACTA back‐propagation neural network was applied to predicting the KIC values using tensile material data and investigating the effects of crack plane orientation and temperature. The 595 KIC data of structural steels were used for training and testing the neural network model. In the trained neural network model, yield stress has relatively the most effect on KIC value among tensile material properties and KIC value was more sensitive to KIC test temperature than to crack plane orientation valid in the range of material data covered in this study. The performance of the trained artificial neural network (ANN) was evaluated by comparing output of the ANN with results of a conventional least squares fit to an assumed shape. The conventional linear or nonlinear least squares fitting methods gave very poor fitting results but the results predicted by the trained neural network were considerably satisfactory. This study shows that the neural network can be a good tool to predict KIC values according to the variation of the temperature and the crack plane orientation using tensile test results.

In this chapter, the primary focus is on theoretical and empirical study of functional link neural networks (FLNNs) for classification. We present a hybrid Chebyshev functional link neural network (cFLNN) without hidden layer with evolvable particle swarm optimization (ePSO) for classification. The resulted classifier is then used for assigning proper class label to an unknown sample. The hybrid cFLNN is a type of feed-forward neural networks, which have the ability to transform the non-linear input space into higher dimensional space where linear separability is possible. In particular, the proposed hybrid cFLNN combines the best attribute of evolvable particle swarm optimization (ePSO), back-propagation learning (BP-Learning), and Chebyshev functional link neural networks (CFLNN). We have shown its effectiveness of classifying the unknown pattern using the datasets obtained from UCI repository. The computational results are then compared with other higher order neural networks (HONNs) like functional link neural network with a generic basis functions, Pi-Sigma neural network (PSNN), radial basis function neural network (RBFNN), and ridge polynomial neural network (RPNN).

Multilayer perceptron (MLP) has been found to be most frequently used model for stock market forecasting. MLP is characterized with black-box in nature and lack of providing a formal method of deriving ultimate structure of the model. Higher order neural network (HONN) has the ability to expand the input representation space, perform high learning capabilities that require less memory in terms of weights and nodes and have been utilized in many complex data mining problems. To capture the extreme volatility, nonlinearity and uncertainty associated with stock data, this paper considered a HONN, called Pi-Sigma Neural Network (PSNN), for prediction of closing prices of five real stock markets. The tunable weights are optimized by Gradient Descent (GD) and a global search technique, Genetic Algorithm (GA). The model proves its superiority when trained with GA in terms of Average Percentage of Errors (APE).

BACKGROUND: Currently, artificial intelligence in the form of artificial neural networks is being actively implemented in a number of areas of our lives, including medicine. In particular, in otorhinolaryngology, artificial neural networks are used to analyze images obtained during endoscopic examinations of patients (e.g., videolaryngoscopy) [1–3]. The interpretation of laryngoscopic images often presents significant difficulties for practicing physicians, which reduces the frequency of detection of precancerous laryngeal diseases and contributes to the increase in the number of patients with stage III–IV laryngeal cancer [4, 5]. This underscores the significance of prompt performance and accurate interpretation of the findings of endoscopic examinations of patients with laryngeal disorders. Artificial neural networks can be employed to analyze the results of videolaryngoscopy, furnishing the physician with supplementary information that can enhance diagnostic accuracy and diminish the probability of error [6, 7]. AIM: The study aims to develop and train an artificial neural network for recognizing characteristic features of laryngeal neoplasms and variants of laryngeal normality. MATERIALS AND METHODS: The study was conducted under the grant of the Moscow Center for Innovative Technologies in Healthcare (grant No. 2112-1/22) entitled “Using Neural Networks (Artificial Intelligence Algorithms) for Control and Improving the Quality of Diagnosis and Treatment of Diseases of Laryngeal and Ear Structures through Digital Technologies”.The following methods were used during the course of the study: data collection for the creation of a photobank (dataset) of medical images obtained during videolaryngoscopy; data partitioning for the formation of datasets for individual nosologies and groups of diseases; the method of consilium; analysis of the accuracy of recognition and classification of digital endoscopic images; and training of classification neural networks. Consequently, a dataset comprising 1,471 laryngeal images in digital formats (JPEG, BMP) was assembled, labelled, and uploaded for the purpose of training the artificial neural network. Of the total number of images, 410 were classified as pertaining to laryngeal formation, while 1061 were classified as variants of normality. Subsequently, the neural network was trained and tested to identify the signs of normal and laryngeal masses. RESULTS: The results of the testing of the artificial neural network indicated the formation of an inaccuracy matrix, the calculation of the value of recognition accuracy, the calculation of the quality indicators of the model performance, and the construction of the ROC curve. The developed and trained artificial neural network demonstrated an accuracy of 86% in recognizing the signs of laryngeal masses and norms. CONCLUSIONS: This study demonstrates that a trained artificial neural network can successfully distinguish between signs of normal and laryngeal masses in endoscopic photographs. With further training of the neural network and achievement of high accuracy, this technology can be used in clinical practice as an assistant in the interpretation of laryngoscopic images and early diagnosis of laryngeal masses. It can also be employed to control and improve the quality of diagnosis and treatment of diseases of the throat, nose, and ears by primary care physicians.

Time series forecasting gets much attention due to its impact on many practical applications. Higher-order neural network with recurrent feedback is a powerful technique which used successfully for forecasting. It maintains fast learning and the ability to learn the dynamics of the series over time. In general, the most used recurrent feedback is the network output. However, no much attention has been paid to use network error instead of the network output. For that, in this paper, we propose a novel model which is called Ridge Polynomial Neural Network with Error Feedback (RPNN-EF) that combines the properties of higher order and error feedback recurrent neural network. Three signals have been used in this paper, namely heat wave temperature, IBM common stock closing price and Mackey–Glass equation. Simulation results show that RPNN-EF is significantly faster than other RPNN-based models for one-step ahead forecasting and its forecasting performance is more significant than these models for multi-step ahead forecasting.

Proposes a fuzzy neural expert system (FNES) which has a feedforward fuzzy neural network whose input layer consists of fuzzy cell groups and crisp (non-fuzzy) cell groups. Here, the truthfulness of fuzzy information and crisp information of training data is represented by fuzzy cell groups and crisp cell groups, respectively. The expert system has the following two functions: generalization of the information derived from the training data and embodiment of knowledge in the form of the fuzzy neural network; and extraction of fuzzy If-Then rules with linguistic relative importance of each proposition in an antecedent (If-part) from a trained fuzzy neural network. The paper also gives a method to extract automatically fuzzy If-Then rules from the trained neural network. To prove the effectiveness and validity of the proposed fuzzy neural expert system, a fuzzy neural expert system for medical diagnosis has been developed. >

Target Recognition Based on Directional Filter Banks and Higher-Order Neural Networks

Traditionally, the application of a neural network (Haykin, 1999) to solve a problem has required to follow some steps before to obtain the desired network. Some of these steps are the data preprocessing, model selection, topology optimization and then the training. It is usual to spend a large amount of computational time and human interaction to perform each task of before and, particularly, in the topology optimization and network training. There have been many proposals to reduce the effort necessary to do these tasks and to provide the experts with a robust methodology. For example, Giles et al. (1995) provides a constructive method to optimize iteratively the topology of a recurrent network. Other methods attempt to reduce the complexity of the network structure by mean of removing unnecessary network nodes and connections like in (Morse, 1994). In the last years, evolutionary algorithms have been shown as promising tools to solve this problem, existing many competitive approaches in the literature. For example, Blanco et al. (2001) proposed a master-slave genetic algorithm to train (master algorithm) and to optimize the size of the network (slave algorithm). For a general view of the problem and the use of evolutionary algorithms for neural network training and optimization, we refer the reader to (Yao, 1999). Although the literature about genetic algorithms and neural networks is very extensive, we would like to remark the recent popularity of multi-objective optimization (Coello et al., 2002, Jin, 2006), specially to solve the problem of simultaneous training and topology optimization of neural networks. These methods have shown to perform suitably for this task in previous works, although most of them are proposed for feedforward models. They attempt to optimize the structure of the network (number of connections, hidden units or layers), while training the network at the same time. Multi-objective algorithms may provide important advantages in the simultaneous training and optimization of neural networks: They may force the search to return a set of optimal networks instead of a single one; they are capable to speed-up the optimization process; they may be preferred to a weight-aggregation procedure to cover the regularization problem in neural networks; and they are more suitable when the designer would like to combine different error measures for the training. A recent review of these techniques may be found in (Jin, 2006).

In 1984 Hopfield showed that the time evolution of a symmetric Hopfield neural networks are a motion in state space that seeks out minima in the energy function (i.e., equilibrium point set of Hopfield neural networks). Because high-order Hopfield neural networks have more extensive applications than Hopfield neural networks, and have been discussed on the convergence of the networks. In practice, a neural network is often subject to environmental noise. It is therefore useful and interesting to find out whether the high-order neural network system still approacher some limit set under stochastic perturbation. In this paper, we will give a number of useful bounds for the noise intensity under which the stochastic high-order neural network will approach its limit set. Our result cancels the requirement of symmetry of the connection weight matrix and includes the classic result on Hopfield neural networks, which is a special case of stochastic high-order Hopfield neural networks. In the end, A example is given to verify the effective of our results.

In this paper, by using the Young inequality technique and introducing many real parameters, some criteria are established for the boundedness and exponential stability of a class of high-order neural networks with time-varying delay. The upper bound of the solutions are also estimated, which is seldom considered before. The results obtained in this paper generalize and improve the existing ones since they can be considered as a special case of our ones as a lower-order case. Illustrative example is also given in the end of this paper to show the effectiveness of our results.

Training of neural network can be accelerated by limited numerical precision together with specialized low-precision hardware. This paper studies how low precision can impact on entire training of RNNs. We emulate low precision training for recently proposed gated recurrent unit (GRU) and use dynamic fixed point as a target numeric format. We first show that batch normalization on input sequences can help speed up training with low precision as well as high precision. We also show that the overflow rate should be carefully controlled for dynamic fixed point. We study low precision training with various rounding options including bit truncation, round to nearest, and stochastic rounding. Stochastic rounding shows superior results than the other options. The effect of fully low precision training is also analyzed by comparing partial low precision training. We show that the piecewise linear activation function with stochastic rounding can achieve comparable training results with floating point precision. Low precision multiplier and accumulator (MAC) with linear-feedback shift register (LFSR) is implemented with 28nm Synopsys PDK for energy and performance analysis. Implementation results show low precision hardware is 4.7x faster, and energy per task is up to 4.55x lower than that of floating point hardware.

There have been various loss functions proposed to improve the training of neural networks with sigmoid activation output nodes. For neural networks with softmax activation output nodes, the cross-entropy loss function is commonly used for training and several attempts have been made to improve the performance of such networks by modifying the standard cross-entropy loss. However, rather than simply aiming to improve overall classification accuracy, it is often necessary to address misclassification costs differently depending on their real-world importance. In practice, the cost of errors can vary greatly across domains such as finance, security, insurance, and healthcare. From this perspective, this paper proposes a modified cross-entropy loss function designed to control the training of neural networks with softmax output nodes. The effectiveness of the proposed loss function is demonstrated through simulations on the CEDAR handwritten digit recognition task, showing that classification performance can be adjusted according to the order of the modified loss function. This approach will serve as the basis for designing a novel cost-sensitive learning method tailored to neural networks with softmax outputs.

In this paper, we present a new key generation model for image hashing using neural network, which does not embed any data into the content but is able to extract meaningful data from target image. This model trains artificial neural network to assign predefined code and uses this trained artificial neural network weight and the coordinates of the selected feature sub blocks of target image as keys to extract the predefined code. In this model, the observed output signal from the trained neural network is used as image hash value which distinguishes the target image from other images. The proposed method contributes to secure image hashing for content identification without damaging or losing any detailed data of visual images. The proposed method realizes an application for image authentication, image similarity comparison, verification of image integrity and copyright protection of multimedia contents.