Removal of bird-contaminated wind profiler data based on neural networks

This paper presents the results of a study that relied on trainable neural network classifiers to identify and remove bird-contaminated data from wind measurements recorded by a 1290-MHz wind profiler. A wind profiler is a Doppler radar system measuring the three-dimensional wind field. Migrating birds crossing the radar beam can lead to erroneous wind observations. Bird removal was performed by training conventional feedforward neural networks (FFNNs) and quantum neural networks (QNNs) to identify and remove bird-contaminated data recorded by a 1290-MHz wind profiler. A series of experiments evaluated several sets of input features extracted from wind profiler data, various FFNNs and QNNs of different sizes, and criteria employed for identifying birds in wind profiler data.     

Recurrent neural network based prediction of epileptic seizures in intra- and extracranial EEG

Predicting the onset of epileptic seizure is an important and difficult biomedical problem, which has attracted substantial attention of the intelligent computing community over the past two decades. We apply recurrent neural networks (RNN) combined with signal wavelet decomposition to the problem. We input raw EEG and its wavelet-decomposed subbands into RNN training/testing, as opposed to specific signal features extracted from EEG. To the best of our knowledge this approach has never been attempted before. The data used included both scalp and intracranial EEG recordings obtained from two epileptic patients. We demonstrate that the existence of a “preictal” stage (immediately preceding seizure) of some minutes duration is quite feasible.     

混合前馈型神经网络在入侵检测中的应用研究

提出一种基于混沌神经元的混合前馈型神经网络,用于检测复杂的网络入侵模式．这种神经网络具有混沌神经元的延时、收集、思维和分类的功能,避免了MLP神经网络仅能识别网络中当前的滥用入侵行为的弱点．对混合网络进行训练后,将该网络用于滥用入侵检测．使用DARPA数据集对该方法进行评估,结果表明该方法可有效地提高对具备延时特性的Probe和DOS入侵的检测性能．     

量子神经元结构设计及其应用

在量子理论的框架内,解释了神经元的信息处理机制,提出了量子神经元．该神经元对信息的处理分为两阶段．第1阶段为宏观信息收集部分,产生控制量子比特;第2阶段为微观信息处理部分,根据控制量子比特,改变神经元的状态．整个过程模拟量子受控非门．采用人工和实际数据集,作为分类研究对象。对比传统的神经元网络,量子神经元网络显示出较好的分类效果．以丙烯腈反应器作为建模研究对象,该网络显示出较强的泛化能力．     

基于改进量子遗传算法的过程神经元网络训练

针对过程神经元网络由于模型参数较多BP算法不易收敛的问题,提出一种基于量子位Bloch坐标的量子遗传算法.将该算法融合于过程神经网络的训练.按权值参数的个数确定量子染色体上的基因数并完成种群编码,通过新的量子旋转门完成个体的更新.算法中的每条染色体携带3条基因链,因此可扩展对解空间的遍历性,加速优化进程.以两组二维三角函数的模式分类问题为例,仿真结果表明该方法不仅收敛速度快,而且寻优能力强.     

An application of pruning in the design of neural networks for real time flood forecasting

We propose the application of pruning in the design of neural networks for hydrological prediction. The basic idea of pruning algorithms, which have not been used in water resources problems yet, is to start from a network which is larger than necessary, and then remove the parameters that are less influential one at a time, designing a much more parameter-parsimonious model. We compare pruned and complete predictors on two quite different Italian catchments. Remarkably, pruned models may provide better generalization than fully connected ones, thus improving the quality of the forecast. Besides the performance issues, pruning is useful to provide evidence of inputs relevance, removing measuring stations identified as redundant (30–40% in our case studies) from the input set. This is a desirable property in the system exercise since data may not be available in extreme situations such as floods; the smaller the set of measuring stations the model depends on, the lower the probability of system downtimes due to missing data. Furthermore, the Authority in charge of the forecast system may decide for real-time operations just to link the gauges of the pruned predictor, thus saving costs considerably, a critical issue in developing countries.

基于小波包分解和遗传神经网络的癫痫识别

基于小波包分解和遗传神经网络对正常脑电和癫痫脑电进行识别。通过分析脑电数据找出信号特征;利用一维离散小波包分解提取含有识别特征的脑电信号频率段,并以脑电各频段的相对能量作为信号特征;然后建立基于遗传算法优化的BP网络,用于对癫痫脑电识别。实验结果表明,该方法可以有效提取信号特征,并且对信号进行准确的识别。     

基于灰色神经网络的入侵检测系统研究

将灰色预测和神经网络有机的结合起来,构造出了新的灰色神经网络GNNM,并用于入侵检测系统(IDS)中,仿真结果表明,GNNM算法在较低误报率的基础上达到了理想的检测率,与传统的神经网络算法相比,不但提高了系统的并行计算能力和系统的可用信息的利用率,还提高了系统的建模效率与模型精度.     

基于集成预测的稀有时间序列检测

为了解决误判问题,从预测的角度给出了离群点的定义,并提出了预测可信度和离群度的概念;同时,提出采用置换技术来降低离群点对预测模型的影响,并提出了基于集成预测的稀有时间序列检测算法。针对真实数据集的实验表明,可信度和离群度的定义是合理的,稀有时间序列检测算法是有效的。     

基于卷积神经网络的目标检测研究综述

随着训练数据的增加以及机器性能的提高,基于卷积神经网络的目标检测冲破了传统目标检测的瓶颈,成为当前目标检测的主流算法。因此,研究如何有效地利用卷积神经网络进行目标检测具有重要的价值。首先回顾了卷积神经网络如何解决传统目标检测中存在的问题;其次介绍了卷积神经网络的基本结构,叙述了当前卷积神经网络的研究进展以及常用的卷积神经网络;然后重点分析和讨论了两种应用卷积神经网络进行目标检测的思路和方法,指出了目前存在的不足;最后总结了基于卷积神经网络的目标检测,以及未来的发展方向。     

Application of radial basis function and feedforward artificial neural networks to the Escherichia coli fermentation process

Radial basis function and feedforward neural networks are considered for modelling of the recombinant Escherichia coli fermentation process. The models use industrial on-line data from the process as input variables in order to estimate the concentrations of biomass and recombinant protein, normally only available from off-line laboratory analysis. The models performances are compared by prediction error and graphical fit using results obtained from a common testing set of fermentation data.     

改进的无线传感网混沌Hopfield盲检测算法

在密集部署的无线传感器网络中,相邻传感器的信号可能高度相关。在无线传感网传输模型分簇的基础上,针对Hopfield神经网络极易陷入局部最优解、收敛速度慢等缺陷,利用混沌序列的遍历性和类随机性,提出一种改进的混沌Hopfield盲检测算法解决无线传感网簇内传感器信号盲检测问题。算法的思想是:利用混沌映射产生初始发送序列,并且在算法出现早熟收敛时进行小幅度的混沌扰动,借此降低算法的误码率。仿真结果表明,改进的混沌Hopfield神经网络算法所需数据量极短,从而成功实现簇内簇首传感器信号盲检测。     

Regularizing the effect of input noise injection in feedforward neural networks training

Injecting input noise during feedforward neural network (NN) training can improve generalization performance markedly. Reported works justify this fact arguing that noise injection is equivalent to a smoothing regularization with the input noise variance playing the role of the regularization parameter. The success of this approach depends on the appropriate choice of the input noise variance. However, it is often not known a priori if the degree of smoothness imposed on the FNN mapping is consistent with the unknown function to be approximated. In order to have a better control over this smoothing effect, a loss function putting in balance the smoothed fitting induced by the noise injection and the precision of approximation, is proposed. The second term, which aims at penalizing the undesirable effect of input noise injection or controlling the deviation of the random perturbed loss, was obtained by expressing a certain distance between the original loss function and its random perturbed version. In fact, this term can be derived in general for parametrical models that satisfy the Lipschitz property. An example is included to illustrate the effectiveness of learning with this proposed loss function when noise injection is used.     

Wavelet based fault detection in analog VLSI circuits using neural networks

This paper deals with a new method of testing analog VLSI circuits, using wavelet transform for analog circuit response analysis and artificial neural networks (ANN) for fault detection. Pseudo-random patterns generated by Linear Feedback Shift Register (LFSR) are used as input test patterns. The wavelet coefficients obtained for the fault-free and faulty cases of the circuits under test (CUT) are used to train the neural network. Two different architectures, back propagation and probabilistic neural networks are trained with the test data. To minimize the neural network architecture, normalization and principal component analysis are done on the input data before it is applied to the neural network. The proposed method is validated with two IEEE benchmark circuits, namely, the operational amplifier and state variable filter.     

An artificial neural network for fault detection in the assembly of thread-forming screws

This paper presents a method, based on a two-layer dynamic Elman neural network, for detecting faults in the assembly of thread-forming screws. Using torque measurements, the method provides a high degree of reliability in detecting assembly faults. The ability of neural networks to learn and to generalize creates an efficient detection system when there is limited or distorted information available about the assembly process.     

An introduction to neural networks and their applications in manufacturing

It is a frequently quoted fact that today's manufacturing functions are becoming more and more inter-disciplinary, with new approaches and techniques continuously and rapidly introduced and adopted. The recent applications of neural networks in manufacturing provide a typical example of this trend. This paper examines the structures and functions of neural networks, and provides some examples of their manufacturing applications.     

基于量子神经网络的启发式网络隐蔽信道检测模型

为了提高隐蔽信道的检测率,讨论了传统的隐蔽信道检测技术的原理并对其特点作了详细的对比研究;通过对网络隐蔽通道特点的归纳和分析,找到可以描述网络隐蔽通道的若干属性,并提出基于网络隐蔽通道特征指纹的检测思想,归纳出隐蔽信道在协议域、统计规律、行为特征等方面表现出的通信指纹信息,在此基础上,设计并实现了一个基于量子神经网络的启发式网络隐蔽信道检测模型。测试结果表明该检测模型运行高效、检测结果较准确。     

基于卷积神经网络的胸片肺结节检测

针对目前胸片的肺结节检测方案的检出率较低,且存在大量的假阳性的问题,提出了一种新的基于卷积神经网络(CNN)的肺结节检测方案.增强肺结节区域的图像信号;选择正、负样本训练卷积神经网络模型,检测结节时用滑动窗口的方法对增强后的图片进行处理得到候选区域;根据候选区域的面积排除假阳性.方案中省略了传统方法中的肺区分割步骤,避免了因此可能丢失的肺结节图像.在日本放射技术学会(JSRT)数据库上测试结果显示,系统在平均每幅图5.0个假阳性水平下敏感度为86％,对不明显和非常不明显的结节检出率达到了84％,优于当前相关文献报道的方法.     

Hybrid system of ART and RBF neural networks for online clustering

An online clustering task is considered for machine state monitoring purpose. In the previous authors’ researches a classical ART-2 network was tested for online classification of operational states in the context of a wind turbine monitoring. Some drawbacks, however, were found when a data stream size had been increased. This case is investigated in this paper. Classical ART-2 network can cluster data incorrectly when data points are compared by using Euclidean distance. Furthermore, ART-2 network can lose accuracy when data stream is processed for long time. The way of improving the ART-2 network is considered and two main steps of that are taken. At first, the stereographic projection is implemented. At the second step, a new type of hybrid neural system which consists of ART-2 and RBF networks with data processed by using the stereographic projection is introduced. Tests contained elementary scenarios for low-dimensional cases as well as higher dimensional real data from wind turbine monitoring. All the tests implied that an efficient system for online clustering had been found.     

3D多尺度深度卷积神经网络肺结节检测

目的 肺结节是肺癌的早期存在形式。低剂量CT（computed tomogragphy）扫描作为肺癌筛查的重要检查手段,已经大规模应用于健康体检,但巨大的CT数据带来了大量工作,随着人工智能技术的快速发展,基于深度学习的计算机辅助肺结节检测引起了关注。由于肺结节尺寸差别较大,在多个尺度上表示特征对结节检测任务至关重要。针对结节尺寸差别较大导致的结节检测困难问题,提出一种基于深度卷积神经网络的胸部CT序列图像3D多尺度肺结节检测方法。方法 包括两阶段：1）尽可能提高敏感度的结节初检网络;2）尽可能减少假阳性结节数量的假阳性降低网络。在结节初检网络中,以组合了压缩激励单元的Res2Net网络为骨干结构,使同一层卷积具有多种感受野,提取肺结节的多尺度特征信息,并使用引入了上下文增强模块和空间注意力模块的区域推荐网络结构,确定候选区域;在由Res2Net网络模块和压缩激励单元组成的假阳性降低网络中对候选结节进一步分类,以降低假阳性,获得最终结果。结果 在公共数据集LUNA16（lung nodule analysis 16）上进行实验,实验结果表明,对于结节初检网络阶段,当平均每例假阳性个数为22时,敏感度可达到0.983,相比基准ResNet + FPN（feature pyramid network）方法,平均敏感度和最高敏感度分别提高了2.6%和0.8%;对于整个3D多尺度肺结节检测网络,当平均每例假阳性个数为1时,敏感度为0.924。结论 与现有主流方案相比,该检测方法不但提高了肺结节检测的敏感度,还有效地控制了假阳性,取得了更优的性能。