Light gradient boosting machine with optimized hyperparameters for identification of malicious access in IoT network

In this paper, an advanced and optimized Light Gradient Boosting Machine (LGBM) technique is proposed to identify the intrusive activities in the Internet of Things (IoT) network. The followings are the major contributions: i) An optimized LGBM model has been developed for the identification of malicious IoT activities in the IoT network; ii) An efficient evolutionary optimization approach has been adopted for finding the optimal set of hyper-parameters of LGBM for the projected problem. Here, a Genetic Algorithm (GA) with k-way tournament selection and uniform crossover operation is used for efficient exploration of hyper-parameter search space; iii) Finally, the performance of the proposed model is evaluated using state-of-the-art ensemble learning and machine learning-based model to achieve overall generalized performance and efficiency. Simulation outcomes reveal that the proposed approach is superior to other considered methods and proves to be a robust approach to intrusion detection in an IoT environment.     

Determination of thallium in water samples

This article illustrates the developments of effective preconcentration techniques and highly sensitive detection methods for accurate measurements of Tl species at extremely low concentration in aqueous solutions. The literature on this topic was taken from the Analytical Abstracts in the period 1990-2005.     

Optimal filter design subject to output sidelobe constraints: Computational algorithm and numerical results

An algorithm is presented for the design of optimal detection filters in radar and communications systems, subject to inequality constraints on the maximum output sidelobe levels. This problem was reduced in an earlier paper (Ref. 1) to an unconstrained one in the dual space of regular Borel measures, with a nondifferentiable cost functional. Here, the dual problem is solved via steepest descent, using the directional Gateaux differential. The algorithm is shown to be convergent, and numerical results are presented.This research was supported by the Australian Research Grants Committee.     

Non-destructive testing of tubes using a time reverse numerical simulation (TRNS) method

A method for the detection of defects in cylindrical structures and the determination of their positions and orientations is presented in this paper. The scattered field, which is generated by the interaction of excited guided waves with a defect, is evaluated with an approach named time reverse numerical simulation method (TRNS). Since the excited waves and the scattered field propagate along the sample, the time-consuming scanning of the whole tube can be eliminated. The scattered displacement field is measured in three dimensions over time with a laser vibrometer at different locations distributed equally around the circumference at a fixed axial coordinate far away from the defect. Instead of analyzing the complicated time signals directly, they are played back in time. If the recorded displacement histories of the scattered field are reversed in time and played back in an identical structure, the waves travel back the same path and interfere to a maximum at their origin. The result is an amplitude increase at the position of the defect where the scattered field was generated. Instead of playing back the recorded time signals in an experiment, this step is replaced by a numerical simulation. Only this enables the visualization and detection of the amplitude increase. As long as the simulation is of high accuracy, the position of the maximum interference corresponds exactly to the location of the defect in the experiment, although no defect is implemented in the simulation.     

Highly Effective Detection of Amitraz in Honey by Using Surface-Enhanced Raman Scattering Spectroscopy Coupled with Chemometric Methods

As an effective and universal acaricide, amitraz is widely used on beehives against varroasis caused by the mite Varroa jacobsoni. Its residues in honey pose a great danger to human health. In this study, a sensitive, rapid, and environmentally friendly surface-enhanced Raman spectroscopy method (SERS) was developed for the determination of trace amount of amitraz in honey with the use of silver nanorod (AgNR) array substrate. The AgNR array substrate fabricated by an oblique angle deposition technique exhibited an excellent SERS activity with an enhancement factor of ∽10⁷. Density function theory was employed to assign the characteristic peak of amitraz. The detection of amitraz was further explored and amitraz in honey at concentrations as low as 0.08 mg/kg can be identified. Specifically, partial least square regression analysis was employed to correlate the SERS spectra in full-wavelength with C_amitraz to afford a multiple-quantitative amitraz predicting model. Preliminary results show that the predicted concentrations of amitraz in honey samples are in good agreement with their real concentrations. Compared with the conventional univariate quantitative model based on single peak’s intensity, the proposed multiple-quantitative predicting model integrates all the characteristic peaks of amitraz, thus offering an improved detecting accuracy and anti-interference ability.     

Ribonucleotide and ribonucleoside determination by ambient pressure ion mobility spectrometry

Detection limits and reduced mobilities for 12 ribonucleotides and 4 ribonucleosides were measured by ambient pressure electrospray ionization-ion mobility spectrometry (ESI-IMS). With the instrument used in this study it was possible to separate some of these compounds within mixtures. Detection limits reported for ribonucleotides and ribonucleosides ranged from 15 to 300 pmol and the reduced mobilities ranged from 41 to 56 suggesting that ambient pressure ESI-IMS may be used for their rapid and sensitive separation and detection. This report demonstrates that it was possible to use ion mobility spectrometry (IMS) to obtain a spectrum for the separation of nucleotides and nucleosides in less than 1 min. The application holds great promise for nucleotide analysis in the area of separating DNA fragments in genome sequencing and also for forensics DNA typing examinations used for the identification of blood stains in crime scenes and paternity testing.     

地下金属管线管径的探测

用现有的一台电磁感应式地下金属管线探测仪在现场进行人工数据采样,然后拟合出一个地下金属单管线二次场分布的经验公式,再用该分布函数曲线半宽度参数,伴以地下全属管线埋深的探测,便可探测出管径值,探测误差约为实际管径的10％。     

基于检测-跟踪联动的紧凑型高频地波雷达弱目标自适应检测方法

紧凑型地波雷达发射功率低、回波信噪比低,目标检测难度大,在目标跟踪时由于目标漏检经常出现航迹断裂。为了提升其目标检测性能,提出了一种基于检测-跟踪联动的弱目标自适应检测方法。当跟踪器检测到目标航迹无法关联到新点迹时,将当前目标预测状态反馈至检测器;检测器在距离-多普勒谱上建立局部检测波门,采用二元假设检验法感知波门内的检测背景,根据不同的检测背景选取适用的检测门限调整方法,降低恒虚警检测的门限,判定是否有弱目标被检出。若能够检出目标,对其进行测向后将新点迹发送至跟踪器进行处理。利用实测数据开展了目标检测与跟踪实验,结果表明：与先检测后跟踪方法相比,该方法得到的目标航迹时长增加了29.76%,平均延长了19.25 min。     

Determination of low analyte concentrations by near-infrared spectroscopy: effect of spectral pretreatments and estimation of multivariate detection limits

Near infrared spectroscopy (NIRS) was used in combination with partial least squares (PLS) calibration to determine low concentrated analytes. The effect of the orthogonal signal correction (OSC) and net analyte signal (NAS) pretreatments on the models obtained at concentrations of analyte near its detection limit was studied. Both pretreatments were found to accurately resolve the analyte signal and allow the construction of PLS models from a reduced number of factors; however, they provided no substantial advantage in terms of %RSE for the prediction samples. Multiple methodologies for the estimation of detection limits could be found in the bibliography. Nevertheless, detection limits were determined by a multivariate method based on the sample-specific standard error for PLS regression, and compared with the univariate method endorsed by ISO 11483. The two methods gave similar results, both being effective for the intended purpose of estimating detection limits for PLS models. Although OSC and NAS allow isolating the analyte signal from the matrix signal, they provide no substantial improvement in terms of detection limits. The proposed method was used to the determine 2-ethylhexanol at concentrations from 20 to 1600 ppm in an industrial ester. The detection limit obtained, round 100 ppm, testifies to the ability of NIR spectroscopy to detect low concentrated analytes.