共识径向基神经网络应用于近红外光谱法测定三七中总黄酮

将共识策略结合径向基神经网络用于近红外光谱法测定三七中总黄酮的含量中。首先采用离散小波变换对近红外光谱进行预处理,去除噪声并压缩数据。继而采用共识径向基神经网络建立校正模型。结果表明:共识策略可以使模型更稳定、更准确。     

非高斯噪声增强纳米尺度钯粒子表面一氧化碳氧化反应随机速率振荡

过去的几十年里,人们对噪声的积极作用进行了广泛研究,其中,化学反应体系中噪声的作用引起人们的很大兴趣.人们研究了单晶表面和纳米粒子表面催化体系中内、外噪声对反应振荡的影响,发现了随机共振和相干共振现象.然而,在以往的研究中,人们总是用高斯噪声代替实际噪声.但实验研究表明,神经元和自催化反应等生物化学反应体系中可能存在非高斯噪声.本文研究非高斯噪声对纳米尺度钯粒子表面一氧化碳催化氧化反应速率振荡的影响.我们首先给出了不存在外噪声时体系振荡随粒子尺度的变化情况,得到了内噪声相干共振行为.其次,我们重点研究了在非高斯噪声作用下体系反应速率的振荡.发现,随着非高斯噪声与高斯噪声的偏离q的增大,振荡规律性经历了由差变好到再变差的演化,而且在某个最佳的q值,振荡变得最有规律性.我们分别给出了振荡时间序列、功率谱和信噪比随q值的变化,充分反映了上述振荡的演化情况.这个结果表明,适当的外噪声能够增强纳米尺度钯粒子表面一氧化碳氧化反应随机速率振荡,而且存在某个最佳的外噪声,使速率振荡得到最大程度的增强.同时表明,非高斯噪声可能比高斯噪声起到更大的增强作用.最后,我们对非高斯噪声增强一氧化碳反应速率振荡这一现象的物理机制作了简单...     

基于导数直方图和神经网络的色谱基线提取算法

根据基线信号的变化特征,提出了一种新的色谱基线提取算法－基于一阶导数直方图和神经网络的色谱基线提取算法。鉴于信号总是存在着噪声,含噪声的基线信号一阶导数值也较大,并且分布无规律,而采用了多分辨率小波变换滤波技术。通过选择阈值门限,本算法可精确提取出信号的基线,尤其适用于非基线对称类的色谱信号的基线（趋势函数）提取。     

主成分-人工神经网络在近红外光谱定量分析中的应用

近红外光谱的主成分由非线性迭代偏最小二乘法（ＮＩＰＡＬＳ）求出。主成分作标准化处理后,作为Ｂ－Ｐ神经网络的输入结点进行非线性迭代。该法的优点是,充分利用了全光谱的数据,得到消除噪声后的最佳主成分,能建立非线性模型,Ｂ－Ｐ神经网络迭代时间显著缩短。用该法对大麦中的淀粉含量进行了定量分析研究。结果为：校准和预测的相关系数分别为０．９８１和０．９５３,校准和预测的相对标准偏差分别为１．７０％和２．４８％。     

窗口因子分析算法的改进及带噪声HPLCDAD数据的解析

窗口因子分析(WFA)是渐进过程二维数据解析的有力工具, 但信号中的噪声对解析结果有较强的干扰作用. 误差分析表明抽象因子分析(AFA)保留的噪声对解析结果有很大的影响. 提出了一种改进的WFA算法, 并对模拟和实验HPLC-DAD数据进行了解析. 结果表明, 改进的算法能够很大程度地克服噪声的影响, 改善由噪声而导致的峰形畸变, 从而提高了WFA处理带噪声数据的能力.     

红敏光致聚合物全息记录材料噪声研究

本文从几个不同的方面研究了明胶全息记录材料体系(PPG)产生噪声的原因,并研究了单体丙烯酰胺(Am)与交联剂亚甲基双丙烯酰胺(B.Am)浓度比值、三乙醇胺(TEA)浓度、曝光量、后处理条件等对噪声的影响.结果表明:单体/交联剂比例为30:1时,体系噪声有所降低.三乙醇胺浓度对体系噪声影响较大,适宜浓度为1%.后处理工艺中,水洗温度对噪声有直接影响,最佳温度为28℃.     

噪声性质对B-Z流动反应体系随机共振影响的模拟

用一个三变量的B-Z流动反应模型研究当流速参数kf在Hopf分岔点附近被色噪声和白噪声调制后系统动力学的行为. 通过计算机模拟观察到噪声诱导的振荡以及随机共振(stochastic resonance)现象, 以及色噪声的相关时间和白噪声的持续时间对随机共振行为的影响.     

二维小波变换与子窗口因子分析法结合用于含噪声HPLC-DAD数据的解析

将二维小波变换与子窗口因子分析法相结合,用于模拟的高噪声HPLC-DAD数据的解析.首先应用二维小波变换滤除噪声,然后采用子窗口因子分析法解析重叠峰.结果表明,信噪比为10的高噪声基本被滤除掉.     

与高斯噪声偏离增强Hodgkin-Huxley神经元激发相干性（摘要）

自从Hodgkin和Huxley对神经元动力学行为进行开创性研究以来,内、外噪声对神经元膜电势激发的积极作用被广泛研究.以往研究中,人们通常认为噪声遵循高斯分布.然而,有关的生物学实验研究表明,在诸如神经元等感官体系中存在非高斯噪声.近十年,一类特殊形式的非高斯噪声在非线性动力学体系（包括生物体系和化学体系）中的作用受到人们的关注.本文中,我们用数值模拟方法研究该类非高斯噪声对确定性Hodgkin-Huxley（HH）神经元激发行为的影响.     

主成分—人工神经网络在近红外光谱定量分析中的应用

近红外光谱的主成分由非线性迭代偏最小二乘法（ＮＩＰＡＬＳ）求出,主成分作标准化处理后,作为Ｂ－Ｐ神经网络的输入结点进行非线性迭代。该法的优点是,充分利用了全光谱的数据,得到消除噪声后的最佳主成分,能建立非线性模型,Ｂ－Ｐ神经网络迭代时间显著缩短,用该法对大麦中的淀粉含量进行了定量分析研究。结果为：校准和预测的相关系数分别为０．９８１和０．９５３,校准和预测的相对标准偏差分别为１．７０％和２．４８％     

Recurrent neural network based hybrid model for reconstructing gene regulatory network

One of the exciting problems in systems biology research is to decipher how genome controls the development of complex biological system. The gene regulatory networks (GRNs) help in the identification of regulatory interactions between genes and offer fruitful information related to functional role of individual gene in a cellular system. Discovering GRNs lead to a wide range of applications, including identification of disease related pathways providing novel tentative drug targets, helps to predict disease response, and also assists in diagnosing various diseases including cancer. Reconstruction of GRNs from available biological data is still an open problem. This paper proposes a recurrent neural network (RNN) based model of GRN, hybridized with generalized extended Kalman filter for weight update in backpropagation through time training algorithm. The RNN is a complex neural network that gives a better settlement between biological closeness and mathematical flexibility to model GRN; and is also able to capture complex, non-linear and dynamic relationships among variables. Gene expression data are inherently noisy and Kalman filter performs well for estimation problem even in noisy data. Hence, we applied non-linear version of Kalman filter, known as generalized extended Kalman filter, for weight update during RNN training. The developed model has been tested on four benchmark networks such as DNA SOS repair network, IRMA network, and two synthetic networks from DREAM Challenge. We performed a comparison of our results with other state-of-the-art techniques which shows superiority of our proposed model. Further, 5% Gaussian noise has been induced in the dataset and result of the proposed model shows negligible effect of noise on results, demonstrating the noise tolerance capability of the model.     

集成滤波器在谱图信号除噪中的应用

本文提出了一种由神经元网络与线性自适应滤波器组成的集成滤波器的实现方法，用于谱图信号除噪处理。     

基于分形维数的分析信号自适应中值滤波

提出一种面向分析仪器谱图信号处理的分形维自适应中值波方法（AMeFFD）。该法延拓运用分形理论，定义了相对点盒维数概念，由此建立建立判定脉冲型噪声的特异性指标，从而可自动调节中值滤波窗口宽度，有效地滤除脉冲型噪声及其它类噪声。对仿真色谱信号及实测色谱图的处理结果表明：AMeFFD法克服了经典中值滤波算法的缺陷，无论在信号的均方根偏差还是谱峰差等指标上，均明显优于后者，能在确保谱峰不畸变的同时更有效地滤除脉冲型常见噪声，是处理化学谱图信号的有力工具。     

A simulation study to evaluate the statistical noise and spatial resolution in image reconstruction of emission computed tomography--with respect to the optimization of the filter function in the convolution integral

Filtered backprojection method has been commonly used to reconstruct images in the field of the computed tomography (CT). However, in the emission CT such as positron and single photon CT, poor counting static which are caused by limited dosage to patients, limited counting rate capacity and limited efficiency of the imaging device, produce a statistical noise in the reconstructed image. The magnitude of the statistical noise and the spatial resolution were evaluated for various shapes of the filter used in the convolution integrals of the filtered back-projection procedure. The statistical noise was proportional to the inverse of the root of the total number of counts for any filters. The high-frequency-cut characteristic of the filter reduced the statistical noise, but increased the spatial resolution in the images. It was possible to optimize the shape of the filter for given total number of counts and required statistical noise and spatial resolution.     

自适应形态滤波器用于色谱信号的处理

建立了一种能自适应地调整尺度的形态滤波器对色谱信号中混杂的离散冲型噪声加以滤除的方法。结果表明,自适应形态滤波器在滤除脉冲型噪声进,比目前文献报道的中值滤波方法呈现出更好的效果,它能保存高度完整的信号信息,是处理色谱信号脉冲型噪声的有力工具。     

Tailoring noise frequency spectrum to improve NIR determinations

Near infrared spectroscopy (NIR) contains excessive background noise and weak analytical signals caused by near infrared overtones and combinations. That makes it difficult to achieve quantitative determinations of low concentration samples by NIR. A simple chemometric approach has been established to modify the noise frequency spectrum to improve NIR determinations. The proposed method is to multiply one Savitzky-Golay filtered NIR spectrum with another reference spectrum added with thermal noises before the other Savitzky-Golay filter. Since Savitzky-Golay filter is a kind of low-pass filter and cannot eliminate low frequency components of NIR spectrum, using one step or two consecutive Savitzky-Golay filter procedures cannot improve the determination of NIR greatly. Meanwhile, significant improvement is achieved via the Savitzky-Golay filtered NIR spectrum processed with the multiplication alteration before the other Savitzky-Golay filter. The frequency range of the modified noise spectrum shifts toward higher frequency regime via multiplication operation. So the second Savitzky-Golay filter is able to provide better filtering efficiency to obtain satisfied result. The improvement of NIR determination with tailoring noise frequency spectrum technique was demonstrated by both simulated dataset and two measured NIR spectral datasets. It is expected that noise frequency spectrum technique will be adopted mostly in applications where quantitative determination of low concentration sample is crucial.     

Quantitative fuzzy neural network for analytical determination

A quantitative fuzzy neural network (Q-FNN) for pattern recognition in analytical determination is reported in this paper. The fuzzy neural network (FNN) combines a fuzzy logic system with an artificial neural network (ANN) so that it has both advantages of a high training speed and strong anti-interference. Importantly, the analytical concept of relative error (RE) in quantitative determination has been integrated into FNN so that the Q-FNN provides a very good quantitative capability in chemical analysis, and prevents the system from an over-fitting problem. The logarithm curve with noise in terms of analytical response versus concentration is calibrated by trained FNN and a close approximation to the ideal one without noise is obtained. The Q-FNN has been applied to the concentration determination of freon in the presence of interference gases. The prediction error for a test set in quantification is less than 10% while no qualitative mistake is observed, implying that the quantitative FNN has sustained the feature of pattern recognition. The results indicate that the Q-FNN has obvious advantages not only in converging speed, but also in the quantitative accuracy over the ANN.     

Improvement of spectral resolution in the presence of periodic noise and microphonics for hyper-pure germanium detector gamma-ray spectrometry using a new digital filter

The radiation dose of workers and patients resulting from inhaling radon and through the consumption of spring waters was examined in the hospital near the Héviz lake in Hungary. The radiation dose originating from radon was 2.15–3.95 mSv·y⁻¹ concerning workers at the spa. The radiation dose originating from radon in the case of those regularly taking a bath was an average of 0.75 mSv·y⁻¹. Due to the limited duration of treatments a bound effective dose of maximum 100 μSv·y⁻¹ may originate from radon and inhaling radon, while a maximum of 1.4 μSv·y⁻¹ may originate from ingestion of ²²²Rn, ²²⁶Ra, ²³⁴U and ²³⁸U radionuclides.     

新极谱法实验数据的卡尔曼滤波

本文提出的"1:n+2"新卡尔曼滤波器,具有算法简单、占用计算机内存少、运行速度快等优点;同时又能有效地消除信号中的随机噪音和电容电流,讨论了"1:3"滤波器的结构性能,采用Cd⁺离子的阶梯扫描伏安法的实测数据来分析滤波结果的最优性,用于处理TI⁺、pb²⁺、In³⁺等离子的实测数据,均获得理想的滤波效果。