A Novel Method for Pattern Recognition of GIS Partial Discharge via Multi-Information Ensemble Learning

Partial discharge (PD) is the main feature that effectively reflects the internal insulation defects of gas-insulated switchgear (GIS). It is of great significance to diagnose the types of insulation faults by recognizing PD to ensure the normal operation of GIS. However, the traditional diagnosis method based on single feature information analysis has a low recognition accuracy of PD, and there are great differences in the diagnosis effect of various insulation defects. To make the most of the rich insulation state information contained in PD, we propose a novel multi-information ensemble learning for PD pattern recognition. First, the ultra-high frequency and ultrasonic data of PD under four typical defects of GIS are obtained through experiment. Then the deep residual convolution neural network is used to automatically extract discriminative features. Finally, multi-information ensemble learning is used to classify PD types at the decision level, which can complement the shortcomings of the independent recognition of the two types of feature information and has higher accuracy and reliability. Experiments show that the accuracy of the proposed method can reach 97.500%, which greatly improves the diagnosis accuracy of various insulation defects.     

Forecasting COVID-19 Epidemic Trends by Combining a Neural Network with Rt Estimation

On 31 December 2019, a cluster of pneumonia cases of unknown etiology was reported in Wuhan (China). The cases were declared to be Coronavirus Disease 2019 (COVID-19) by the World Health Organization (WHO). COVID-19 has been defined as SARS Coronavirus 2 (SARS-CoV-2). Some countries, e.g., Italy, France, and the United Kingdom (UK), have been subjected to frequent restrictions for preventing the spread of infection, contrary to other ones, e.g., the United States of America (USA) and Sweden. The restrictions afflicted the evolution of trends with several perturbations that destabilized its normal evolution. Globally, Rt has been used to estimate time-varying reproduction numbers during epidemics. Methods: This paper presents a solution based on Deep Learning (DL) for the analysis and forecasting of epidemic trends in new positive cases of SARS-CoV-2 (COVID-19). It combined a neural network (NN) and an Rt estimation by adjusting the data produced by the output layer of the NN on the related Rt estimation. Results: Tests were performed on datasets related to the following countries: Italy, the USA, France, the UK, and Sweden. Positive case registration was retrieved between 24 February 2020 and 11 January 2022. Tests performed on the Italian dataset showed that our solution reduced the Mean Absolute Percentage Error (MAPE) by 28.44%, 39.36%, 22.96%, 17.93%, 28.10%, and 24.50% compared to other ones with the same configuration but that were based on the LSTM, GRU, RNN, ARIMA (1,0,3), and ARIMA (7,2,4) models, or an NN without applying the Rt as a corrective index. It also reduced MAPE by 17.93%, the Mean Absolute Error (MAE) by 34.37%, and the Root Mean Square Error (RMSE) by 43.76% compared to the same model without the adjustment performed by the Rt. Furthermore, it allowed an average MAPE reduction of 5.37%, 63.10%, 17.84%, and 14.91% on the datasets related to the USA, France, the UK, and Sweden, respectively.     

Harnessing Protein-Ligand Interaction Fingerprints to Predict New Scaffolds of RIPK1 Inhibitors

Necroptosis has emerged as an exciting target in oncological, inflammatory, neurodegenerative, and autoimmune diseases, in addition to acute ischemic injuries. It is known to play a role in innate immune response, as well as in antiviral cellular response. Here we devised a concerted in silico and experimental framework to identify novel RIPK1 inhibitors, a key necroptosis factor. We propose the first in silico model for the prediction of new RIPK1 inhibitor scaffolds by combining docking and machine learning methodologies. Through the data analysis of patterns in docking results, we derived two rules, where rule #1 consisted of a four-residue signature filter, and rule #2 consisted of a six-residue similarity filter based on docking calculations. These were used in consensus with a machine learning QSAR model from data collated from ChEMBL, the literature, in patents, and from PubChem data. The models allowed for good prediction of actives of >90, 92, and 96.4% precision, respectively. As a proof-of-concept, we selected 50 compounds from the ChemBridge database, using a consensus of both molecular docking and machine learning methods, and tested them in a phenotypic necroptosis assay and a biochemical RIPK1 inhibition assay. A total of 7 of the 47 tested compounds demonstrated around 20–25% inhibition of RIPK1’s kinase activity but, more importantly, these compounds were discovered to occupy new areas of chemical space. Although no strong actives were found, they could be candidates for further optimization, particularly because they have new scaffolds. In conclusion, this screening method may prove valuable for future screening efforts as it allows for the exploration of new areas of the chemical space in a very fast and inexpensive manner, therefore providing efficient starting points amenable to further hit-optimization campaigns.     

基于注意力机制与迁移学习的乳腺钼靶肿块分类

针对乳腺钼靶图像中良恶性肿块难以诊断的问题,提出一种基于注意力机制与迁移学习的乳腺钼靶肿块分类方法,并用于医学影像中乳腺钼靶肿块的良恶性分类.首先,构建一种新的网络模型,该模型将注意力机制CBAM(Convolutional Block Attention Module)与残差网络ResNet50相结合,用于提高网络对...     

基于机器学习的直接电离质谱爆炸物检测方法

直接电离质谱系统在现场快速检测中的应用日益广泛,主要用于爆炸物、毒品、食品添加剂等的检测.然而,直接电离质谱系统中质谱信号波动大且同一浓度样品峰强呈现对数正态分布,严重影响了检出限附近低浓度样品的检测准确性.该研究将乙酰水杨酸(115个样品)作为爆炸物模拟物,利用介质阻挡放电离子源与质谱系统,研究了基于机器学习的直接电...     

早期糖尿病风险预测模型的比较研究

糖尿病是一种比较常见的慢性疾病,并且存在较长的无症状阶段.本文主要介绍了机器学习中的5种分类算法,分别是朴素贝叶斯、支持向量机、逻辑回归、决策树和集成分类器Random Forest,并在Weka数据挖掘平台上,对糖尿病数据进行挖掘分析,根据混淆矩阵、Kappa系数、ROC曲线、均方根误差以及相对绝对误差这几个性能指标...     

Multivariate Time Series Imputation: An Approach Based on Dictionary Learning

The problem addressed by dictionary learning (DL) is the representation of data as a sparse linear combination of columns of a matrix called dictionary. Both the dictionary and the sparse representations are learned from the data. We show how DL can be employed in the imputation of multivariate time series. We use a structured dictionary, which is comprised of one block for each time series and a common block for all the time series. The size of each block and the sparsity level of the representation are selected by using information theoretic criteria. The objective function used in learning is designed to minimize either the sum of the squared errors or the sum of the magnitudes of the errors. We propose dimensionality reduction techniques for the case of high-dimensional time series. For demonstrating how the new algorithms can be used in practical applications, we conduct a large set of experiments on five real-life data sets. The missing data (MD) are simulated according to various scenarios where both the percentage of MD and the length of the sequences of MD are considered. This allows us to identify the situations in which the novel DL-based methods are superior to the existing methods.     

Stochastic Control for Bayesian Neural Network Training

In this paper, we propose to leverage the Bayesian uncertainty information encoded in parameter distributions to inform the learning procedure for Bayesian models. We derive a first principle stochastic differential equation for the training dynamics of the mean and uncertainty parameter in the variational distributions. On the basis of the derived Bayesian stochastic differential equation, we apply the methodology of stochastic optimal control on the variational parameters to obtain individually controlled learning rates. We show that the resulting optimizer, StochControlSGD, is significantly more robust to large learning rates and can adaptively and individually control the learning rates of the variational parameters. The evolution of the control suggests separate and distinct dynamical behaviours in the training regimes for the mean and uncertainty parameters in Bayesian neural networks.     

Adversarial Machine Learning for NextG Covert Communications Using Multiple Antennas

This paper studies the privacy of wireless communications from an eavesdropper that employs a deep learning (DL) classifier to detect transmissions of interest. There exists one transmitter that transmits to its receiver in the presence of an eavesdropper. In the meantime, a cooperative jammer (CJ) with multiple antennas transmits carefully crafted adversarial perturbations over the air to fool the eavesdropper into classifying the received superposition of signals as noise. While generating the adversarial perturbation at the CJ, multiple antennas are utilized to improve the attack performance in terms of fooling the eavesdropper. Two main points are considered while exploiting the multiple antennas at the adversary, namely the power allocation among antennas and the utilization of channel diversity. To limit the impact on the bit error rate (BER) at the receiver, the CJ puts an upper bound on the strength of the perturbation signal. Performance results show that this adversarial perturbation causes the eavesdropper to misclassify the received signals as noise with a high probability while increasing the BER at the legitimate receiver only slightly. Furthermore, the adversarial perturbation is shown to become more effective when multiple antennas are utilized.     

基于深度学习的ELM实时识别研究

基于深度学习的方法,在HL-2A装置上开发出了一套边缘局域模(ELM)实时识别算法。算法使用5200次放电数据(约24.19万数据切片)进行学习,得到一个深度为22层的卷积神经网络。为衡量算法的识别能力,识别了HL-2A装置自2009年实现稳定ELMyH模放电以来所有历史数据(约26000次放电数据),共识别出1665次H模放电,其中误识别35次,误报率为2.10%。在实际的1634次H模放电中,漏识别4次,漏识别率为0.24%。该误报率和漏报率可以满足ELM实时识别的精度要求。识别算法在实时控制环境下,对单个时间点的平均计算时间为0.46ms,可以满足实时控制的计算速度要求。