Interval type-2 fuzzy logic based antenatal care system using phonocardiography

The assessment of fetal wellbeing depends heavily on variations in fetal heart rate (FHR) patterns. The variations in FHR patterns are very complex in nature thus its reliable interpretation is very difficult and often leads to erroneous diagnosis. We propose a new method for evaluation of fetal health status based on interval type-2 fuzzy logic through fetal phonocardiography (fPCG). Type-2 fuzzy logic is a powerful tool in handling uncertainties due to extraneous variations in FHR patterns through its increased fuzziness of relations. Four FHR parameters are extracted from each fPCG signal for diagnostic decision making. The membership functions of these four inputs and one output are chosen as a range of values so as to represent the level of uncertainty. The fuzzy rules are constructed based on standard clinical guidelines on FHR parameters. Experimental clinical tests have shown very good performance of the developed system in comparison with the FHR trace simultaneously recorded through standard fetal monitor. Statistical evaluation of the developed system shows 92% accuracy. With the proposed method we hope that, long-term and continuous antenatal care will become easy, cost effective, reliable and efficient.     

Detection of Spikes with Artificial Neural Networks Using Raw EEG

Artificial neural networks (ANN) using raw electroencephalogram (EEG) data were developed and tested off-line to detect transient epileptiform discharges (spike and spike/wave) and EMG activity in an ongoing EEG. In the present study, a feedforward ANN with a variable number of input and hidden layer units and two output units was used to optimize the detection system. The ANN system was trained and tested with the backpropagation algorithm using a large data set of exemplars. The effects of different EEG time windows and the number of hidden layer neurons were examined using rigorous statistical tests for optimum detection sensitivity and selectivity. The best ANN configuration occurred with an input time window of 150 msec (30 input units) and six hidden layer neurons. This input interval contained information on the wave component of the epileptiform discharge which improved detection. Two-dimensional receiver operating curves were developed to define the optimum threshold parameters for best detection. Comparison with previous networks using raw EEG showed improvement in both sensitivity and selectivity. This study showed that raw EEG can be successfully used to train ANNs to detect epileptogenic discharges with a high success rate without resorting to experimenter-selected parameters which may limit the efficiency of the system.     

Stock market prediction using artificial neural networks with optimal feature transformation

This paper compares a feature transformation method using a genetic algorithm (GA) with two conventional methods for artificial neural networks (ANNs). In this study, the GA is incorporated to improve the learning and generalizability of ANNs for stock market prediction. Daily predictions are conducted and prediction accuracy is measured. In this study, three feature transformation methods for ANNs are compared. Comparison of the results achieved by a feature transformation method using the GA to the other two feature transformation methods shows that the performance of the proposed model is better. Experimental results show that the proposed approach reduces the dimensionality of the feature space and decreases irrelevant factors for stock market prediction.     

基于特征融合卷积神经网络的端到端加密流量分类

针对现有人工神经网络方法在网络加密流量分类应用中结构复杂且计算量大的问题,首次提出了一种基于特征融合的轻量级网络模型Inception-CNN,用于端到端加密流量的分类,在显著提高分类结果准确性的同时,大大降低了网络计算复杂度。利用Inception模块1×1卷积进行降维,减少了计算参数;从不同的感受野中做到不同级别上的特征提取,将多种不同尺寸滤波器卷积的特征进行融合,从而在原始数据中提取到更加丰富的特征自动学习原始输入和预期输出之间的非线性关系;利用池化操作没有参数的特性,防止产生过拟合。选择使用国际公开ISCX VPN-nonVPN数据集作为实验数据,采用softmax作为分类器,实现了对加密流量的准确分类。实验结果表明,该模型分类准确率达到97.3%、精确率达到97.2%、召回率达到97.7%、F1-score达到97.5%,并且对不同类别的加密流量识别效果也更加均衡。     

Analysis of impact of balanced level on MI-based and non-MI-based feature selection methods

Advancements in high-speed computer technology play an ever-increasing role in analyzing various types and massive size data. However, handling big high-dimensional data sets is a challenge in terms of computational storage and capacity. Through feature selection methods, data dimensions can be reduced by eliminating the dummy variables, allowing for more extensive analysis. In this paper, data are classified based on the ratio of response into two types: balanced (almost the same ratio for each class) and partially balanced (consists of majority and minority, with virtually the same ratio for minority classes). Performance comparisons of various feature selection methods for balanced and partially balanced data are provided. This approach will help in selecting sampling strategy and feature selection methods that perform well while utilizing appropriate resources for high-dimensional data.

     

Neural Network and Classification Approach in Identifying Customer Behavior in the Banking Sector: A Case Study of an International Bank

The customer relationship focus for banks is in development of main competencies and strategies of building strong profitable customer relationships through considering and managing the customer impression, influence on the culture of the bank, satisfactory treatment, and assessment of valued relationship building. Artificial neural networks (ANNs) are used after data segmentation and classification, where the designed model register records into two class sets, that is, the training and testing sets. ANN predicts new customer behavior from previously observed customer behavior after executing the process of learning from existing data. This article proposes an ANN model, which is developed using a six‐step procedure. The back‐propagation algorithm is used to train the ANN by adjusting its weights to minimize the difference between the current ANN output and the desired output. An evaluation process is conducted to determine whether the ANN has learned how to perform. The training process is halted periodically, and its performance is tested until an acceptable result is obtained. The principles underlying detection software are grounded in classical statistical decision theory.     

C-Net: A Method for Generating Non-deterministic and Dynamic Multivariate Decision Trees

Despite the fact that artificial neural networks (ANNs) are universal function approximators, their black box nature (that is, their lack of direct interpretability or expressive power) limits their utility. In contrast, univariate decision trees (UDTs) have expressive power, although usually they are not as accurate as ANNs. We propose an improvement, C-Net, for both the expressiveness of ANNs and the accuracy of UDTs by consolidating both technologies for generating multivariate decision trees (MDTs). In addition, we introduce a new concept, recurrent decision trees, where C-Net uses recurrent neural networks to generate an MDT with a recurrent feature. That is, a memory is associated with each node in the tree with a recursive condition which replaces the conventional linear one. Furthermore, we show empirically that, in our test cases, our proposed method achieves a balance of comprehensibility and accuracy intermediate between ANNs and UDTs. MDTs are found to be intermediate since they are more expressive than ANNs and more accurate than UDTs. Moreover, in all cases MDTs are more compact (i.e., smaller tree size) than UDTs. Received 27 January 2000 / Revised 30 May 2000 / Accepted in revised form 30 October 2000     

A time-multiplexed reconfigurable neuroprocessor

Built around biological-like information-processing structures, artificial neural networks (ANNs) have demonstrated their power and usefulness in areas where identification and adaptability are most crucial. After sufficient training with a given set of problem data (using an arbitrary learning rule), ANNs can independently form internal representations (models) of the data's underlying phenomenon. ANNs typically have a large number of highly interconnected processing elements, which constrains their hardware implementation and network architecture. To obtain high density and processing-element connectivity, most ANN architectures employ some kind of resource sharing. A multilayer structure allows processing elements to share communication lines and control circuitry, so we chose to develop a multilayered neuroprocessor based on pipelined time-step interneural communication. Other pulse stream architectures use the rate of impulses to represent the neural states. Our architecture, however, represents the neural states through pulse amplitude modulation. Also, in our design, the analog computation consists of integrating the bipolar input pulses corresponding to the excitatory and inhibitory activations. The processing-element analog path consists. of CMOS transmission gates controlled by buffered signals originating from the neuroprocessor control unit. The processing elements broadcast their output states, held on a local capacitor, during their assigned time slots. These features are desirable to meet the design goals of versatility, high density, high connectivity, and scalability     

Efficient wavelet-based ECG processing for single-lead FHR extraction

This paper describes a novel model for fetal heart rate (FHR) monitoring from single-lead mother?s abdomen ECG (AECG) measurements. This novel method is divided in two stages: the first step consists on a one-step wavelet-based preprocessing for simultaneous baseline and high-frequency noise suppression, while the second stage efficiently detects fetal QRS complexes allowing FHR monitoring. The presented structure has been simplified as much as possible, in order to reduce computational cost and thus enable possible custom hardware implementations. Moreover, the proposed scheme and its fixed-point modeling have been tested using real abdominal ECG signals, which allow the validation of the presented approach and provide high accuracy.     

A Two Stage Sequential Ensemble Applied to the Classification of Alzheimer’s Disease Based on MRI Features

We present a two stage sequential ensemble where data samples whose output from the first classifier fall in a low confidence output interval (LCOI) are processed by a second stage classifier. Training is composed of three processes: training the first classifier, determining the LCOI of the first classifier, and training the second classifier upon the data items whose output fall in the LCOI. The LCOI is determined varying a threshold on the false positive rate (FPR) and false negative rate (FNR) curves. We have tested the approach on a database of feature vectors for the classification of Alzheimer’s disease (AD) and control subjects extracted from structural magnetic resonance imaging (sMRI) data. In this paper, we focus on the combinations obtained when the first classifier is a relevance vector machine (RVM). Obtained results improve over previous results for this database.     

CHORD RECOGNITION USING NEURAL NETWORKS BASED ON PARTICLE SWARM OPTIMIZATION

A sequence of musical chords can facilitate musicians in music arrangement and accompaniment. To implement an intelligent system for chord recognition, in this article we propose a novel approach using artificial neural networks (ANN) trained bythe particle swarm optimization (PSO) technique and back-propagation (BP) learning algorithm. All of the training and testing data are generated from musical instrument digital interface (MIDI) symbolic data. Furthermore, in order to improve the recognition efficiency, an additional feature of cadencesis included. In other words, cadence is not only the structural punctuation of a melodic phrase but is considered as the important feature for chord recognition. Experimental results of our proposed approach show that adding a cadence feature significantly improves recognition rate, and the ANN-PSO method outperforms ANN-BP in chord recognition. In addition, because preliminary experimental recognition rates are generally not stable enough, we chose the optimal ANNs to propose a two-phase ANN model to integrate the results among many models.     

Prediction of fetal state from the cardiotocogram recordings using adaptive neuro-fuzzy inference systems

In this study, a new scheme was presented for the prediction of fetal state from fetal heart rate (FHR) and the uterine contraction (UC) signals obtained from cardiotocogram (CTG) recordings. CTG recordings are widely used in pregnancy and provide very valuable information regarding fetal well-being. The information effectively extracted from these recordings can be used to predict pathological state of the fetus and makes an early intervention possible before there is an irreversible damage to the fetus. The proposed scheme is based on adaptive neuro-fuzzy inference systems (ANFIS). Using features extracted from the FHR and UC signals, an ANFIS was trained to predict the normal and the pathological state. The method was tested with clinical data that consist of 1,831 CTG recordings. Out of these 1,831 recordings, 1,655 of them were classified as normal and the remaining 176 were classified as pathological by a consensus of three expert obstetricians. It was demonstrated that the ANFIS-based method was able to classify the normal and the pathologic states with 97.2 and 96.6 % accuracy, respectively.     

Incremental communication for multilayer neural networks: erroranalysis

Artificial neural networks (ANNs) involve a large amount of internode communications. To reduce the communication cost as well as the time of learning process in ANNs, we earlier proposed (1995) an incremental internode communication method. In the incremental communication method, instead of communicating the full magnitude of the output value of a node, only the increment or decrement to its previous value is sent to a communication link. In this paper, the effects of the limited precision incremental communication method on the convergence behavior and performance of multilayer neural networks are investigated. The nonlinear aspects of representing the incremental values with reduced (limited) precision for the commonly used error backpropagation training algorithm are analyzed. It is shown that the nonlinear effect of small perturbations in the input(s)/output of a node does not cause instability. The analysis is supported by simulation studies of two problems. The simulation results demonstrate that the limited precision errors are bounded and do not seriously affect the convergence of multilayer neural networks.     

Use of particle swarm optimization for machinery fault detection

A study is presented on the application of particle swarm optimization (PSO) combined with other computational intelligence (CI) techniques for bearing fault detection in machines. The performance of two CI based classifiers, namely, artificial neural networks (ANNs) and support vector machines (SVMs) are compared. The time domain vibration signals of a rotating machine with normal and defective bearings are processed for feature extraction. The extracted features from original and preprocessed signals are used as inputs to the classifiers for detection of machine condition. The classifier parameters, e.g., the number of nodes in the hidden layer for ANNs and the kernel parameters for SVMs are selected along with input features using PSO algorithms. The classifiers are trained with a subset of the experimental data for known machine conditions and are tested using the remaining set of data. The procedure is illustrated using the experimental vibration data of a rotating machine. The roles of the number of features, PSO parameters and CI classifiers on the detection success are investigated. Results are compared with other techniques such as genetic algorithm (GA) and principal component analysis (PCA). The PSO based approach gave a test classification success rate of 98.6–100% which were comparable with GA and much better than with PCA. The results show the effectiveness of the selected features and the classifiers in the detection of the machine condition.     

非接触式胎儿心率检测

胎儿心率检测是围产期常规检测,是评估孕妇和胎儿健康的主要生理指标.相对现有的接触式胎心检测技术,本文提出一种更为便捷,成本低廉的非接触式胎儿心率提取算法.首先基于欧拉视频颜色放大技术,对视频中颜色信号放大.其次,利用光电容积脉搏波描记法提取血液容积脉冲信号,并对母体噪声进行分离,计算功率谱密度提取.将采集到的胎心率,与医院专用胎心设备检测的结果进行定量分析,数据表明可以达到96%的准确度.     

Fuzzy learning vector quantization for hyperspectral coastal vegetation classification

Artificial neural networks (ANNs) may be of significant value in extracting vegetation type information in complex vegetation mapping problems, particularly in coastal wetland environments. Unsupervised, self-organizing ANNs have not been employed as frequently as supervised ANNs for vegetation mapping tasks, and further remote sensing research involving fuzzy ANNs is also needed. In this research, the utility of a fuzzy unsupervised ANN, specifically a fuzzy learning vector quantization (FLVQ) ANN, was investigated in the context of hyperspectral AVIRIS image classification. One key feature of the neural approach is that unlike conventional hyperspectral data processing methods, endmembers for a given scene, which can be difficult to determine with confidence, are not required for neural analysis. The classification accuracy of FLVQ was comparable to a conventional supervised multi-layer perceptron, trained with backpropagation (MLP) (KHAT (Kˆ) accuracy: 82.82% and 84.66%, respectively; normalized accuracy: 74.60% and 75.85%, respectively), with no significant difference at the 95% confidence level. All neural algorithms in the experiment yielded significantly higher classification accuracies than the conventional endmember-based hyperspectral mapping method assessed (i.e., matched filtering, where Kˆ accuracy = 61.00% and normalized accuracy = 57.96%). FLVQ was also dramatically more computationally efficient than the baseline supervised and unsupervised ANN algorithms tested, including the MLP and the Kohonen self-organizing map (SOM), respectively. The 400-neuron FLVQ network required only 3.6% of the computation time used by the MLP network, and only 5.9% of the MLP time was used by the 588-neuron FLVQ network. In addition, the 400-neuron FLVQ used only 16.7% of the time used by the 400-neuron SOM for model development.     

A subspace approach to balanced truncation for model reduction of nonlinear control systems

In this paper, we introduce a new method of model reduction for nonlinear control systems. Our approach is to construct an approximately balanced realization. The method requires only standard matrix computations, and we show that when it is applied to linear systems it results in the usual balanced truncation. For nonlinear systems, the method makes use of data from either simulation or experiment to identify the dynamics relevant to the input–output map of the system. An important feature of this approach is that the resulting reduced‐order model is nonlinear, and has inputs and outputs suitable for control. We perform an example reduction for a nonlinear mechanical system. Copyright © 2002 John Wiley & Sons, Ltd.     

An advanced method in fetal phonocardiography

The long-term variability of the fetal heart rate (FHR) provides valuable information on the fetal health status. The routine clinical FHR measurements are usually carried out by the means of ultrasound cardiography. Although the frequent FHR monitoring is recommendable, the high quality ultrasound devices are so expensive that they are not available for home care use. The passive and fully non-invasive acoustic recording called phonocardiography, provides an alternative low-cost measurement method. Unfortunately, the acoustic signal recorded on the maternal abdominal surface is heavily loaded by noise, thus the determination of the FHR raises serious signal processing issues. The development of an accurate and robust fetal phonocardiograph has been since long researched. This paper presents a novel two-channel phonocardiographic device and an advanced signal processing method for determination of the FHR. The developed system provided 83% accuracy compared to the simultaneously recorded reference ultrasound measurements.     

Calibrating artificial neural networks by global optimization

Artificial neural networks (ANNs) are used extensively to model unknown or unspecified functional relationships between the input and output of a “black box” system. In order to apply the generic ANN concept to actual system model fitting problems, a key requirement is the training of the chosen (postulated) ANN structure. Such training serves to select the ANN parameters in order to minimize the discrepancy between modeled system output and the training set of observations. We consider the parameterization of ANNs as a potentially multi-modal optimization problem, and then introduce a corresponding global optimization (GO) framework. The practical viability of the GO based ANN training approach is illustrated by finding close numerical approximations of one-dimensional, yet visibly challenging functions. For this purpose, we have implemented a flexible ANN framework and an easily expandable set of test functions in the technical computing system Mathematica. The MathOptimizer Professional global-local optimization software has been used to solve the induced (multi-dimensional) ANN calibration problems.