Genetic algorithms for the optimization of piecewise linear discriminants

The application of genetic algorithms (GAs) to the optimization of piecewise linear discriminants is described. Piecewise linear discriminant analysis (PLDA) is a supervised pattern recognition technique employed in this work for the automated classification of Fourier transform infrared (FTIR) remote sensing data. PLDA employs multiple linear discriminants to approximate a nonlinear separating surface between data categories defined in a vector space. The key to the successful implementation of PLDA is the positioning of the individual discriminants that comprise the piecewise linear discriminant. For the remote sensing application, the discriminant optimization is challenging due to the large number of input variables required and the corresponding tendency for local optima to occur on the response surface of the optimization. In this work, three implementations of GAs are configured and evaluated: a binary-coded GA (GAB), a real-coded GA (GAR), and a Simplex-GA hybrid (SGA). GA configurations are developed by use of experimental design studies, and piecewise linear discriminants for acetone, methanol, and sulfur hexafluoride are optimized (trained). The training and prediction classification results indicate that GAs area viable approach for discriminant optimization. On average, the best piecewise linear discriminant optimized by a GA is observed to classify 11% more analyte-active patterns correctly in prediction than an unoptimized piecewise linear discriminant. Discriminant optimization problems not used in the experimental design study are employed to test the stability of the GA configurations. For these cases, the best piecewise linear discriminant optimized by SGA is shown to classify 19% more analyte-active patterns correctly in prediction than an unoptimized discriminant. These results also demonstrate that the two real number coded GAs (GAR and SGA) perform better than the GAB. Real number coded GAs are also observed to execute faster and are simpler to implement.     

Remote detection of heated ethanol plumes by airborne passive Fourier transform infrared spectrometry

Methodology is developed for the automated detection of heated plumes of ethanol vapor with airborne passive Fourier transform infrared spectrometry. Positioned in a fixed-wing aircraft in a downward-looking mode, the spectrometer is used to detect ground sources of ethanol vapor from an altitude of 2000-3000 ft. Challenges to the use of this approach for the routine detection of chemical plumes include (1) the presence of a constantly changing background radiance as the aircraft flies, (2) the cost and complexity of collecting the data needed to train the classification algorithms used in implementing the plume detection, and (3) the need for rapid interferogram scans to minimize the ground area viewed per scan. To address these challenges, this work couples a novel ground-based data collection and training protocol with the use of signal processing and pattern recognition methods based on short sections of the interferogram data collected by the spectrometer. In the data collection, heated plumes of ethanol vapor are released from a portable emission stack and viewed by the spectrometer from ground level against a synthetic background designed to simulate a terrestrial radiance source. Classifiers trained with these data are subsequently tested with airborne data collected over a period of 2.5 years. Two classifier architectures are compared in this work: support vector machines (SVM) and piecewise linear discriminant analysis (PLDA). When applied to the airborne test data, the SVM classifiers perform best, failing to detect ethanol in only 8% of the cases in which it is present. False detections occur at a rate of less than 0.5%. The classifier performs well in spite of differences between the backgrounds associated with the ground-based and airborne data collections and the instrumental drift arising from the long time span of the data collection. Further improvements in classification performance are judged to require increased sophistication in the ground-based data collection in order to provide a better match to the infrared backgrounds observed from the air.     

Detection and classification of buried dielectric anomalies usingneural networks-further results

The development of a neural network-based detection and classification system for use with buried dielectric anomalies is the main focus of this paper. Several methods of data representation are developed to study their effects on the trainability and generalization capabilities of the neural networks. The method of Karhonen-Loeve (KL) transform is used to extract energy dependent features and to reduce the dimensionality of the weight space of the original data set. To extract the shape-dependent features of the data, another data preprocessing method known as Zernike moments is also studied for its use in the detector/classifier system. The effects of different neural network paradigms, architectural variations, and selection of proper training data on detection and classification rates are studied. Simulation results for nylon and wood targets indicate superior performance when compared to conventional schemes     

Classification of Fourier transform infrared microscopic imaging data of human breast cells by cluster analysis and artificial neural networks

Cluster analysis and artificial neural networks (ANNs) are applied to the automated assessment of disease state in Fourier transform infrared microscopic imaging measurements of normal and carcinomatous immortalized human breast cell lines. K-means clustering is used to implement an automated algorithm for the assignment of pixels in the image to cell and non-cell categories. Cell pixels are subsequently classified into carcinoma and normal categories through the use of a feed-forward ANN computed with the Broyden-Fletcher-Goldfarb-Shanno training algorithm. Inputs to the ANN consist of principal component scores computed from Fourier filtered absorbance data. A grid search optimization procedure is used to identify the optimal network architecture and filter frequency response. Data from three images corresponding to normal cells, carcinoma cells, and a mixture of normal and carcinoma cells are used to build and test the classification methodology. A successful classifier is developed through this work, although differences in the spectral backgrounds between the three images are observed to complicate the classification problem. The robustness of the final classifier is improved through the use of a rejection threshold procedure to prevent classification of outlying pixels.     

Detection and classification of buried dielectric anomalies bymeans of the bispectrum method and neural networks

The development of a neural network-based system for detection and classification of buried landmines is the main focus of this paper. Shape-dependent features are extracted by means of the bispectrum method. These features are then applied to the neural network. A multilayer back-propagation-type neural network is trained and tested on the feature sets extracted from equally spaced radial slices of image windows. Simulation results obtained for two types of targets indicated good detection and classification rates     

DDoS Attack Detection via Multi-Scale Convolutional Neural Network

Distributed Denial-of-Service (DDoS) has caused great damage to the network in the big data environment. Existing methods are characterized by low computational efficiency, high false alarm rate and high false alarm rate. In this paper, we propose a DDoS attack detection method based on network flow grayscale matrix feature via multiscale convolutional neural network (CNN). According to the different characteristics of the attack flow and the normal flow in the IP protocol, the seven-tuple is defined to describe the network flow characteristics and converted into a grayscale feature by binary. Based on the network flow grayscale matrix feature (GMF), the convolution kernel of different spatial scales is used to improve the accuracy of feature segmentation, global features and local features of the network flow are extracted. A DDoS attack classifier based on multi-scale convolution neural network is constructed. Experiments show that compared with correlation methods, this method can improve the robustness of the classifier, reduce the false alarm rate and the missing alarm rate.     

Predicting motor vehicle collisions using Bayesian neural network models: an empirical analysis

Statistical models have frequently been used in highway safety studies. They can be utilized for various purposes, including establishing relationships between variables, screening covariates and predicting values. Generalized linear models (GLM) and hierarchical Bayes models (HBM) have been the most common types of model favored by transportation safety analysts. Over the last few years, researchers have proposed the back-propagation neural network (BPNN) model for modeling the phenomenon under study. Compared to GLMs and HBMs, BPNNs have received much less attention in highway safety modeling. The reasons are attributed to the complexity for estimating this kind of model as well as the problem related to "over-fitting" the data. To circumvent the latter problem, some statisticians have proposed the use of Bayesian neural network (BNN) models. These models have been shown to perform better than BPNN models while at the same time reducing the difficulty associated with over-fitting the data. The objective of this study is to evaluate the application of BNN models for predicting motor vehicle crashes. To accomplish this objective, a series of models was estimated using data collected on rural frontage roads in Texas. Three types of models were compared: BPNN, BNN and the negative binomial (NB) regression models. The results of this study show that in general both types of neural network models perform better than the NB regression model in terms of data prediction. Although the BPNN model can occasionally provide better or approximately equivalent prediction performance compared to the BNN model, in most cases its prediction performance is worse than the BNN model. In addition, the data fitting performance of the BPNN model is consistently worse than the BNN model, which suggests that the BNN model has better generalization abilities than the BPNN model and can effectively alleviate the over-fitting problem without significantly compromising the nonlinear approximation ability. The results also show that BNNs could be used for other useful analyses in highway safety, including the development of accident modification factors and for improving the prediction capabilities for evaluating different highway design alternatives.     

旋转机械故障诊断的神经网络方法研究

BP神经网络具有较好的非线性映射能力,可以描述频率特征和故障之间的关系,而概率神经网络学习规则简单、训练速度快、避免局部极小和反复训练的问题。根据两种神经网络的原理选择合适的参数建立两个旋转机械故障诊断模型,并利用模型对某旋转机械的故障数据进行处理,结果显示两种网络在故障诊断方面的实用价值。通过对故障数据的结果对比可以看到PNN网络比BP网络具有更好的容错能力。     

用概率社会网络进行结构损伤位置识别

在不计测量误差情况下,神经网络能够成功地识别损伤位置及其程度,但在测量噪声影响下,神经网络的损伤识别效果则比较差,考虑到基于多变量模式分类的概率神经网络具有处理受噪声污染的测试数据的能力,本文将可能的损伤位置作为模式类,利用概率神经网络的分类能力来识别结构的损,地对两个算例,一个六层框架和一个两层框架进行数值模拟分析,并将概率神经网络与BP网络进行了比较,结果表明,概率神经网络具有更好的识别效果,是一种很有潜力的结构损伤位置识别方法。     

基于卷积神经网络的光学遥感图像船只检测

目的研究无需进行复杂的图像预处理和人工特征提取,就能提高光学遥感图像的船只检测准确率和实现船只类型精细分类。方法对输入的检测图像,采用选择性搜索的方法产生船只候选区域,用已经标记好的训练样本对卷积神经网络进行监督训练,得到网络参数,然后使用经过监督训练的卷积神经网络提取抽象特征,并对候选区域进行分类,根据船只候选区域的分类概率同时确定船只的位置以及类型。结果与现有的2种检测方法进行对比,实验结果表明卷积神经网络能有效提高船只检测准确率,平均检测准确率达到了93.3%。结论该检测方法无需进行复杂的预处理,能同时对船只进行检测和分类,并能有效提高船只检测准确率。     

Calibration transfer algorithm for automated qualitative analysis by passive Fourier transform infrared spectrometry

The automated qualitative analysis of passive Fourier transform infrared (FT-IR) remote sensing data is made difficult by the presence in the data of background and instrument-specific variation. For data collected with a single instrument, variation in the data arises from changes in the infrared background radiance, changes in the atmospheric composition within the field-of-view of the spectrometer, and changes in the instrument response function arising from temperature variation in the spectrometer. When more than one spectrometer is used, the variation in detector responses and phase signatures between instruments serves to complicate further the task of implementing an automated processing algorithm for detecting the signature of a target compound. In this work, a combination of signal processing and pattern recognition methodology is applied directly to the interferogram data collected by the FT-IR spectrometer to implement an automated compound detection procedure that is independent of background and instrument-specific variation. The key to this algorithm is the use of highly attenuating digital filters to isolate in the interferogram the frequencies associated with an analyte absorption or emission band while suppressing information at other frequencies. For the test compounds, acetone and sulfur hexafluoride, it is demonstrated that when this digital filtering procedure is coupled with either piecewise linear discriminant analysis or a back-propagation neural network, an automated detection algorithm can be developed with data from a primary instrument and then subsequently used to predict the presence of analyte signatures in data collected with a secondary spectrometer. Correct classification rates in excess of 92% are obtained for both compounds when the algorithm is applied to data collected with the secondary instrument.     

Meta-heuristics for Feature Selection and Classification in Diagnostic BreastCancer

One of the most common kinds of cancer is breast cancer. The early detection of it may help lower its overall rates of mortality. In this paper, we robustly propose a novel approach for detecting and classifying breast cancer regions in thermal images. The proposed approach starts with data preprocessing the input images and segmenting the significant regions of interest. In addition, to properly train the machine learning models, data augmentation is applied to increase the number of segmented regions using various scaling ratios. On the other hand, to extract the relevant features from the breast cancer cases, a set of deep neural networks (VGGNet, ResNet-50, AlexNet, and GoogLeNet) are employed. The resulting set of features is processed using the binary dipper throated algorithm to select the most effective features that can realize high classification accuracy. The selected features are used to train a neural network to finally classify the thermal images of breast cancer. To achieve accurate classification, the parameters of the employed neural network are optimized using the continuous dipper throated optimization algorithm. Experimental results show the effectiveness of the proposed approach in classifying the breast cancer cases when compared to other recent approaches in the literature. Moreover, several experiments were conducted to compare the performance of the proposed approach with the other approaches. The results of these experiments emphasized the superiority of the proposed approach.     

An Optimized Approach to Vehicle-Type Classification Using a Convolutional Neural Network

Vehicle type classification is considered a central part of an intelligent traffic system. In recent years, deep learning had a vital role in object detection in many computer vision tasks. To learn high-level deep features and semantics, deep learning offers powerful tools to address problems in traditional architectures of handcrafted feature-extraction techniques. Unlike other algorithms using handcrated visual features, convolutional neural network is able to automatically learn good features of vehicle type classification. This study develops an optimized automatic surveillance and auditing system to detect and classify vehicles of different categories. Transfer learning is used to quickly learn the features by recording a small number of training images from vehicle frontal view images. The proposed system employs extensive data-augmentation techniques for effective training while avoiding the problem of data shortage. In order to capture rich and discriminative information of vehicles, the convolutional neural network is fine-tuned for the classification of vehicle types using the augmented data. The network extracts the feature maps from the entire dataset and generates a label for each object (vehicle) in an image, which can help in vehicle-type detection and classification. Experimental results on a public dataset and our own dataset demonstrated that the proposed method is quite effective in detection and classification of different types of vehicles. The experimental results show that the proposed model achieves 96.04% accuracy on vehicle type classification.     

Digital filtering implementations for the detection of broad spectral features by direct analysis of passive Fourier transform infrared interferograms

Finite impulse response (FIR) filters and finite impulse response matrix (FIRM) filters are evaluated for use in the detection of volatile organic compounds with wide spectral bands by direct analysis of interferogram data obtained from passive Fourier transform infrared (FT-IR) measurements. Short segments of filtered interferogram points are classified by support vector machines (SVMs) to implement the automated detection of heated plumes of the target analyte, ethanol. The interferograms employed in this study were acquired with a downward-looking passive FT-IR spectrometer mounted on a fixed-wing aircraft. Classifiers are trained with data collected on the ground and subsequently used for the airborne detection. The success of the automated detection depends on the effective removal of background contributions from the interferogram segments. Removing the background signature is complicated when the analyte spectral bands are broad because there is significant overlap between the interferogram representations of the analyte and background. Methods to implement the FIR and FIRM filters while excluding background contributions are explored in this work. When properly optimized, both filtering procedures provide satisfactory classification results for the airborne data. Missed detection rates of 8% or smaller for ethanol and false positive rates of at most 0.8% are realized. The optimization of filter design parameters, the starting interferogram point for filtering, and the length of the interferogram segments used in the pattern recognition is discussed.     

Modular artificial neural network for prediction of petrophysicalproperties from well log data

An application of Kohonen's self-organizing map (SOM), learning-vector quantization (LVQ) algorithms, and commonly used backpropagation neural network (BPNN) to predict petrophysical properties obtained from well-log data are presented. A modular, artificial neural network (ANN) comprising a complex network made up from a number of subnetworks is introduced. In this approach, the SOM algorithm is applied first to classify the well-log data into a predefined number of classes, This gives an indication of the lithology in the well. The classes obtained from SOM are then appended back to the training input logs for the training of supervised LVQ. After training, LVQ can be used to classify any unknown input logs. A set of BPNN that corresponds to different classes is then trained. Once the network is trained, it is then used as the classification and prediction model for subsequent input data. Results obtained from example studies using the proposed method have shown to be fast and accurate as compared to a single BPNN network     

Spectral peak verification and recognition using a multilayered neural network

The verification and recognition of peak-shaped signals in analytical data are ubiquitous scientific problems. Experimental data contain overlapping signals and noise, which make sensitive and reliable peak recognition difficult. A peak detection system based on a class of neural networks known as "multilayered perceptrons" has been created. The network was trained and evaluated with use of vapor-phase infrared spectral data. The results of varying the network architecture on system training and prediction performance along with refinement of the form of the input pattern are presented.     

Supervised target detection and classification by training on augmented reality data

A proof of concept for a model-less target detection and classification system for side-scan imagery is presented. The system is based on a supervised approach that uses augmented reality (AR) images for training computer added detection and classification (CAD/CAC) algorithms, which are then deployed on real data. The algorithms are able to generalise and detect real targets when trained on AR ones, with performances comparable with the state-of-the-art in CAD/CAC. To illustrate the approach, the focus is on one specific algorithm, which uses Bayesian decision and the novel, purpose-designed central filter feature extractors. Depending on how the training database is partitioned, the algorithm can be used either for detection or classification. Performance figures for these two modes of operation are presented, both for synthetic and real targets. Typical results show a detection rate of more that 95% and a false alarm rate of less than 5%. The proposed supervised approach can be directly applied to train and evaluate other learning algorithms and data representations. In fact, a most important aspect is that it enables the use of a wealth of legacy pattern recognition algorithms for the sonar CAD/CAC applications of target detection and target classification     

Biomedical Event Extraction Using a New Error Detection Learning Approach Based on Neural Network

Supervised machine learning approaches are effective in text mining, but their success relies heavily on manually annotated corpora. However, there are limited numbers of annotated biomedical event corpora, and the available datasets contain insufficient examples for training classifiers; the common cure is to seek large amounts of training samples from unlabeled data, but such data sets often contain many mislabeled samples, which will degrade the performance of classifiers. Therefore, this study proposes a novel error data detection approach suitable for reducing noise in unlabeled biomedical event data. First, we construct the mislabeled dataset through error data analysis with the development dataset. The sample pairs’ vector representations are then obtained by the means of sequence patterns and the joint model of convolutional neural network and long short-term memory recurrent neural network. Following this, the sample identification strategy is proposed, using error detection based on pair representation for unlabeled data. With the latter, the selected samples are added to enrich the training dataset and improve the classification performance. In the BioNLP Shared Task GENIA, the experiments results indicate that the proposed approach is competent in extract the biomedical event from biomedical literature. Our approach can effectively filter some noisy examples and build a satisfactory prediction model.     

卷积神经网络在印刷品缺陷检测的应用

目的 为了改善传统机器检测印刷产品缺陷存在误费率高的不足。方法 提出以卷积神经网络为控制核心的印刷品缺陷检测系统。设计可在实际检测中应用的卷积神经网络,设计在线印刷质量检测系统的硬件结构。结果 对结构相同而训练次数、学习率不同的卷积神经网络进行了缺陷检测的性能对比,验证了该卷积神经网络在学习率小于0.01时,可以获得较好的识别效果;在学习率大于0.05时,网络不容易收敛。网络训练次数越多,精度越高,相应的训练时间也较长。在满足快速性和精确度的条件下,确定了适应某印刷品的缺陷检验网络训练次数为50,学习率为0.005,此时的识别率为90%。结论 经过实验证明,该检测系统具有良好的缺陷识别能力,缺陷类型的分类准确率较高。该系统具有一定的实用价值。     

Multicontext 3D residual CNN for false positive reduction of pulmonary nodule detection

False positive reduction is a key procedure of computer-aided pulmonary nodule detection. The goal is to recognize the true pulmonary nodule from the plentiful candidates, which received from the first step of pulmonary nodule candidate detection. Convolutional networks can be used to perform false positive nodule reduction, but the classification accuracy need to be further improved. Recently, residual network is more and more popular around the world with its distinguished performance. A multicontext three-dimensional residual convolutional neural network (3D Res-CNN) was proposed to realize the reduction of the false positive nodule. Using two scales of network to adapt the variation of pulmonary nodule size, instead of using an unreferenced function with reference to the identity mapping, 3D Res-CNN uses a shortcut connection to realize the residual structure. For alleviating the data imbalance, firstly patches are rotated and resampled in original images; secondly weights are allotted for different labels in the calculation of cost function. Experiments on volumetric computed tomography (CT) data indicate that our method gets state of the art performance: 0.843 average sensitivity with 0.125, 0.25, 0.5, 1, 2, 4, and 8 false positive per subject. The results show the effectiveness of residual convolutional network for the recognition of the true pulmonary nodule from the plentiful candidates.