Three-dimensional medical image analysis of the heart by the revised GMDH-type neural network self-selecting optimum neural network architecture

In this study, a revised group method of data handling (GMDH)-type neural network algorithm which self-selects the optimum neural network architecture is applied to 3-dimensional medical image analysis of the heart. The GMDH-type neural network can automatically organize the neural network architecture by using the heuristic self-organization method, which is the basic theory of the GMDH algorism. The heuristic self-organization method is a kind of evolutionary computation method. In this revised GMDH-type neural network algorithm, the optimum neural network architecture was automatically organized using the polynomial and sigmoid function neurons. Furthermore, the structural parameters, such as the number of layers, the number of neurons in the hidden layers, and the useful input variables, are selected automatically in order to minimize the prediction error criterion, defined as the prediction sum of squares (PSS).     

Deep multi-layered GMDH-type neural network using revised heuristic self-organization and its application to medical image diagnosis of liver cancer

In this study, the deep multi-layered group method of data handling (GMDH)-type neural network algorithm using revised heuristic self-organization method is proposed and applied to medical image diagnosis of liver cancer. The deep GMDH-type neural network can automatically organize the deep neural network architecture which has many hidden layers. The structural parameters such as the number of hidden layers, the number of neurons in hidden layers and useful input variables are automatically selected to minimize prediction error criterion defined as Akaike’s information criterion (AIC) or prediction sum of squares (PSS). The architecture of the deep neural network is automatically organized using the revised heuristic self-organization method which is a type of the evolutionary computation. This new neural network algorithm is applied to the medical image diagnosis of the liver cancer and the recognition results are compared with the conventional 3-layered sigmoid function neural network.     

Deep feedback GMDH-type neural network and its application to medical image analysis of MRI brain images

The deep feedback group method of data handling (GMDH)-type neural network is applied to the medical image analysis of MRI brain images. In this algorithm, the complexity of the neural network is increased gradually using the feedback loop calculations. The deep neural network architecture is automatically organized so as to fit the complexity of the medical images using the prediction error criterion defined as Akaike’s information criterion (AIC) or prediction sum of squares (PSS). The recognition results show that the deep feedback GMDH-type neural network algorithm is useful for the medical image analysis of MRI brain images, because the optimum neural network architectures fitting the complexity of the medical images are automatically organized so as to minimize the prediction error criterion defined as AIC or PSS.     

Medical image diagnosis of lung cancer by multi-layered GMDH-type neural network self-selecting functions

In this study, a revised Group Method of Data Handling (GMDH)-type neural network self-selecting functions is applied to the computer aided image diagnosis (CAD) of lung cancer. The GMDH-type neural network algorithm has an ability of self-selecting optimum neural network architecture from three neural network architectures, such as sigmoid function neural network, radial basis function neural network and polynomial neural network. The GMDH-type neural network also has abilities of self-selecting the number of layers, the number of neurons in hidden layers and useful input variables. This algorithm is applied to CAD of lung cancers, and it is shown that this algorithm is useful for the CAD, and is very easy to apply to practical complex problems because optimum neural network architecture is automatically organized.     

Feedback GMDH-type neural network algorithm and its application to medical image analysis of cancer of the liver

A revised group method of data handling (GMDH)-type neural network algorithm for medical image recognition is proposed, and is applied to medical image analysis of cancer of the liver. The revised GMDH-type neural network algorithm has a feedback loop and can identify the characteristics of the medical images accurately using feedback-loop calculations. In this algorithm, the polynomial type and the radial basis function (RBF)-type neurons are used for organizing the neural network architecture. The optimum neural network architecture fitting the complexity of the medical images is automatically organized so as to minimize the prediction error criterion, defined as the prediction sum of squares (PSS).     

Medical image diagnosis of kidney regions by deep feedback GMDH-type neural network using principal component-regression analysis

The deep feedback Group Method of Data Handling (GMDH)-type neural network is applied to the medical image recognition of kidney regions. In this algorithm, the principal component-regression analysis is used for the learning calculation of the neural network, and the accurate and stable predicted values are obtained. The neural network architecture is automatically organized so as to fit the complexity of the medical images using the prediction error criterion defined as Akaike’s Information Criterion (AIC) or Prediction Sum of Squares (PSS). The recognition results show that the deep feedback GMDH-type neural network algorithm is useful for the medical image recognition of kidney regions, because the optimum neural network architecture is automatically organized.     

Medical image analysis of abdominal X-ray CT images by deep multi-layered GMDH-type neural network

In this study, a deep multi-layered group method of data handling (GMDH)-type neural network is applied to the medical image analysis of the abdominal X-ray computed tomography (CT) images. The deep neural network architecture which has many hidden layers are automatically organized using the deep multi-layered GMDH-type neural network algorithm so as to minimize the prediction error criterion defined as Akaike’s information criterion (AIC) or prediction sum of squares (PSS). The characteristics of the medical images are very complex and therefore the deep neural network architecture is very useful for the medical image diagnosis and medical image recognition. In this study, it is shown that this deep multi-layered GMDH-type neural network is useful for the medical image analysis of abdominal X-ray CT images.     

Medical image diagnosis of lung cancer by a revised GMDH-type neural network using various kinds of neuron

A revised group method of data handling (GMDH)-type neural network algorithm using various kinds of neuron is applied to the medical image diagnosis of lung cancer. The optimum neural network architecture for medical image diagnosis is automatically organized using a revised GMDH-type neural network algorithm, and the regions of lung cancer are recognized and extracted accurately. In this revised GMDH-type neural network algorithm, polynomial-type and radial basis function (RBF)-type neurons are used for organizing the neural network architecture in order to fit the complexity of the nonlinear system.     

Cascade neural network for binary mapping

The problem of choosing a suitable number of neurons for a neural network which realizes any given binary mapping is automatically solved by the proposed cascade architecture. The utilized algorithm, based on linear programming, the complexity of the resulting net, and its generalization capability are discussed.     

基于自组织递归模糊神经网络的PM2.5浓度预测

针对PM2.5浓度非线性动态变化的特点,提出了一种自组织递归模糊神经网络(self-organizing recurrent fuzzy neural network,SORFNN)方法预测PM2.5小时浓度。首先,通过分析影响PM2.5浓度的多种因素,利用主成分分析法(principal component analysis,PCA)筛选出与PM2.5浓度相关性较强的特征变量作为神经网络的输入变量。然后,根据ε准则和偏最小二乘算法(partial least squares,PLS)进行规则化层神经元的增删,实现递归模糊神经网络结构的自动调整,并采用学习率自适应的梯度下降算法调整模型中心、宽度和权值等参数,建立PM2.5预测模型。最后,利用典型非线性系统辨识和实际PM2.5浓度预测实验进行验证。实验结果表明,所设计的自组织递归模糊神经网络结构精简且预测精度高,较好地满足了PM2.5实时预测的要求。     

输入层自构造神经网络及其在光谱分析中的应用

为了解决红外光谱定量分析中的特征提取和校正规模问题,提出了一种输入层自构造神经网络。这种网络能够利用训练数据的某些先验知识,自动选择输入层神经元的个数。在学习过程中,输入神经元个数从最小值1开始,根据网络误差的变化逐步增加,最终确定最佳神经元数量。这种网络模型将特征提取和参数学习过程融为一体,有利于提高建模效率。利用仿真红外光谱的定量分析实验表明,这种网络模型不仅能够对光谱数据实现高效率的波长选择,并具有抑制随机噪声和非线性干扰的能力。     

A novel model by evolving partially connected neural network for stock price trend forecasting

This paper proposes a novel model by evolving partially connected neural networks (EPCNNs) to predict the stock price trend using technical indicators as inputs. The proposed architecture has provided some new features different from the features of artificial neural networks: (1) connection between neurons is random; (2) there can be more than one hidden layer; (3) evolutionary algorithm is employed to improve the learning algorithm and training weights. In order to improve the expressive ability of neural networks, EPCNN utilizes random connection between neurons and more hidden layers to learn the knowledge stored within the historic time series data. The genetically evolved weights mitigate the well-known limitations of gradient descent algorithm. In addition, the activation function is defined using sin(x) function instead of sigmoid function. Three experiments were conducted which are explained as follows. In the first experiment, we compared the predicted value of the trained EPCNN model with the actual value to evaluate the prediction accuracy of the model. Second experiment studied the over fitting problem which occurred in neural network training by taking different number of neurons and layers. The third experiment compared the performance of the proposed EPCNN model with other models like BPN, TSK fuzzy system, multiple regression analysis and showed that EPCNN can provide a very accurate prediction of the stock price index for most of the data. Therefore, it is a very promising tool in forecasting of the financial time series data.     

一种功能分区的BP神经网络结构设计方法

针对全连接BP网络在解决大规模复杂问题时存在的收敛速度缓慢等问题,提出一种功能分区的BP网络结构模式.利用RBF神经元的物理特性对输入样本空间进行分解,并将分解后的样本送给不同的子BP网络学习.与全连接BP网络相比,降低了网络在学习过程中的权值搜索空间,提高了学习速度,改善了网络泛化性能,体现了人脑在学习过程中的知识积累特征.对三维墨西哥草帽函数逼近和双螺旋分类的实验结果表明,该网络能够解决全连接BP网络不能有效解决的问题.     

Binary object extraction using bi-directional self-organizing neural network (BDSONN) architecture with fuzzy context sensitive thresholding

A novel neural network architecture suitable for image processing applications and comprising three interconnected fuzzy layers of neurons and devoid of any back-propagation algorithm for weight adjustment is proposed in this article. The fuzzy layers of neurons represent the fuzzy membership information of the image scene to be processed. One of the fuzzy layers of neurons acts as an input layer of the network. The two remaining layers viz. the intermediate layer and the output layer are counter-propagating fuzzy layers of neurons. These layers are meant for processing the input image information available from the input layer. The constituent neurons within each layer of the network architecture are fully connected to each other. The intermediate layer neurons are also connected to the corresponding neurons and to a set of neighbors in the input layer. The neurons at the intermediate layer and the output layer are also connected to each other and to the respective neighbors of the corresponding other layer following a neighborhood based connectivity. The proposed architecture uses fuzzy membership based weight assignment and subsequent updating procedure. Some fuzzy cardinality based image context sensitive information are used for deciding the thresholding capabilities of the network. The network self organizes the input image information by counter-propagation of the fuzzy network states between the intermediate and the output layers of the network. The attainment of stability of the fuzzy neighborhood hostility measures at the output layer of the network or the corresponding fuzzy entropy measures determine the convergence of the network operation. An application of the proposed architecture for the extraction of binary objects from various degrees of noisy backgrounds is demonstrated using a synthetic and a real life image.

Prediction of service life of large centrifugal compressor remanufactured impeller based on clustering rough set and fuzzy Bandelet neural network

In order to predict the service life of large centrifugal compressor impeller correctly, the rough set and fuzzy Bandelet neural network are combined to construct the novel prediction model which can give full play to theirs advantages. The attribute reduction algorithm based rough set and clustering method is firstly designed to optimize the inputting variables of fuzzy Bandelet neural network. And then the prediction model based on fuzzy Bandelet neural network is proposed, the Bandelet function is used as the excitation function of hidden layer and is combined with fuzzy theory to improve the prediction effectiveness of the prediction model. The training algorithm of fuzzy Bandelet neural network is designed based on improved genetic algorithm, the improved genetic algorithm introduces the adaptive differential evolution method into the traditional genetic algorithm, which can effectively optimize the parameters of fuzzy Bandelet neural network. Finally, the original 30 input variables of fuzzy Bandelet neural network are reduced to 9 input nodes based on rough set using 500 remanufacturing impellers as research objects. The service life of remanufacturing impeller is predicted based on three prediction models, and simulation results show that the fuzzy Bandelet neural network optimized by improved genetic algorithm has highest prediction precision and efficiency, which can correctly predict the service life of remanufacturing impeller.     

A quantum bi-directional self-organizing neural network (QBDSONN) architecture for binary object extraction from a noisy perspective

This article proposes an efficient technique for binary object extraction in real time from noisy background using quantum bi-directional self-organizing neural network (QBDSONN) architecture. QBDSONN exploits the power of quantum computation. It is composed of three second order neighborhood topology based inter-connected layers of neurons (represented by qubits) arranged as input, intermediate and output layers. In the suggested network architecture, the inter-connection weights and activation values are represented by rotation gates. A self-supervised learning algorithm, suggested in this proposed architecture, relies on the steepest descent algorithm. The quantum neurons enjoy full-connectivity in each layer of the network architecture. The image pixels in terms of qubits are self-organized in between the intermediate or hidden and output layers of the QBDSONN architecture using counter-propagation of the quantum states to obviate time consuming quantum back propagation algorithm. In the final phase, quantum measurement is carried out at the output layer to eliminate superposition of the quantum states of the outputs. In order to establish the result, the proposed QBDSONN architecture is applied on an artificial synthetic and on a real life spanner image with different degrees of uniform and Gaussian noises. Experimental results show that QBDSONN outperforms both its classical counterpart and the supervised auto-associative Hopfield network as far as extraction time is concerned and it retains the shapes of the extracted images with great precision. Experiments are also carried out using a linear method named local statistics (Wiener filter) and a nonlinear technique named median filter with adaptive discrete wavelet transformations (DWT) for binary object extraction to show the dominance of the proposed QBDSONN with respect to the quality of extracted images. Finally, a statistical significance of the proposed QBDSONN is reported by applying 2 sample one sided Kolmogorov–Smirnov test with the existing methods.     

Nonlinear system identification using optimized dynamic neural network

In this paper, both off-line architecture optimization and on-line adaptation have been developed for a dynamic neural network (DNN) in nonlinear system identification. In the off-line architecture optimization, a new effective encoding scheme—Direct Matrix Mapping Encoding (DMME) method is proposed to represent the structure of neural network by establishing connection matrices. A series of GA operations are applied to the connection matrices to find the optimal number of neurons on each hidden layer and interconnection between two neighboring layers of DNN. The hybrid training is adopted to evolve the architecture, and to tune the weights and input delays of DNN by combining GA with the modified adaptation laws. The modified adaptation laws are subsequently used to tune the input time delays, weights and linear parameters in the optimized DNN-based model in on-line nonlinear system identification. The effectiveness of the architecture optimization and adaptation is extensively tested by means of two nonlinear system identification examples.     

基于蚁群算法优化回声状态网络的研究

针对输出权值采用最小二乘法的回声状态网络(ESN),在随机选取输入权值和隐层神经元阈值时,存在收敛速度慢、预测精度不稳定等问题,提出了基于蚁群算法优化回声状态网络(ACO-ESN)的算法。该算法将优化回声状态网络的初始输入权值、隐层神经元阈值问题转化为蚁群算法中蚂蚁寻找最佳路径的问题,输出权值采用最小二乘法计算,通过蚁群算法的更新、变异、遗传等操作训练回声状态网络,选择出使回声状态网络预测误差最小的输入权值和阈值,从而提高其预测性能。将ACO-ESN与ELM、I-ELM、OS-ELM、B-ELM等神经网络的仿真结果进行对比,结果验证经过蚁群算法优化的回声状态网络加快了其收敛速度,改善了其预测性能,并增强了隐层神经元的敏感度。     

Binary image denoising using a quantum multilayer self organizing neural network

Several classical techniques have evolved over the years for the purpose of denoising binary images. But the main disadvantages of these classical techniques lie in that an a priori information regarding the noise characteristics is required during the extraction process. Among the intelligent techniques in vogue, the multilayer self organizing neural network (MLSONN) architecture is suitable for binary image preprocessing tasks.In this article, we propose a quantum version of the MLSONN architecture. Similar to the MLSONN architecture, the proposed quantum multilayer self organizing neural network (QMLSONN) architecture comprises three processing layers viz., input, hidden and output layers. The different layers contains qubit based neurons. Single qubit rotation gates are designated as the network layer interconnection weights. A quantum measurement at the output layer destroys the quantum states of the processed information thereby inducing incorporation of linear indices of fuzziness as the network system errors used to adjust network interconnection weights through a quantum backpropagation algorithm.Results of application of the proposed QMLSONN are demonstrated on a synthetic and a real life binary image with varying degrees of Gaussian and uniform noise. A comparative study with the results obtained with the MLSONN architecture and the supervised Hopfield network reveals that the QMLSONN outperforms the MLSONN and the Hopfield network in terms of the computation time.     

Determining significance of input neurons for probabilistic neural network by sensitivity analysis procedure

In classical feedforward neural networks such as multilayer perceptron, radial basis function network, or counter‐propagation network, the neurons in the input layer correspond to features of the training patterns. The number of these features may be large, and their meaningfulness can be various. Therefore, the selection of appropriate input neurons should be regarded. The aim of this paper is to present a complete step‐by‐step algorithm for determining the significance of particular input neurons of the probabilistic neural network (PNN). It is based on the sensitivity analysis procedure applied to a trained PNN. The proposed algorithm is utilized in the task of reduction of the input layer of the considered network, which is achieved by removing appropriately indicated features from the data set. For comparison purposes, the PNN's input neuron significance is established by using the ReliefF and variable importance procedures that provide the relevance of the input features in the data set. The performance of the reduced PNN is verified against a full structure network in classification problems using real benchmark data sets from an available machine learning repository. The achieved results are also referred to the ones attained by entropy‐based algorithms. The prediction ability expressed in terms of misclassifications is obtained by means of a 10‐fold cross‐validation procedure. Received outcomes point out interesting properties of the proposed algorithm. It is shown that the efficiency determined by all tested reduction methods is comparable.