基于VMD和BSA-KELM的高陡边坡位移预测模型研究

边坡位移的时间序列曲线存在复杂的非线性特性,传统的预测模型精度不足以满足预测要求。为此提出了基于变分模态分解的鸟群优化-核极限学习机的预测模型,并用于河北省某水泥厂的边坡位移预测。该方法首先采用VMD把边坡位移序列分解为一系列的有限带宽的子序列,再对各子序列分别采用相空间重构并用核极限学习机预测,采用鸟群算法优化相空间重构的嵌入维度和KELM中惩罚系数和核参数三个数值,以取得最优预测模型。最后将各个子序列预测值叠加,得到边坡位移的最终预测值。结果表明：和KELM、BSA-KELM、EEMD-BSA-KELM模型相比,基于VMD的BSA-KELM预测精度更高,为边坡位移的预测提供一种有效的方法。     

Inhibitory effects of curcumin and piperine on fluorescent advanced glycation end products formation in a bovine serum albumin–fructose model

This study investigated the inhibitory effects of curcumin and piperine on fluorescent advanced glycation end products (fAGEs) formation in a bovine serum albumin (BSA)–fructose model. Model systems of BSA and fructose were prepared, and curcumin or piperine was added. fAGEs and BSA oxidation product (dityrosine, kynurenine and N'-formylkynurenine) contents were determined. The results showed that fAGEs content decreased with increasing concentration of curcumin and piperine (P < 0.05). Addition of curcumin and piperine at 160 µg mL⁻¹ could inhibit fluorescent AGEs by 100% and 93% respectively. Dityrosine and N'-formylkynurenine contents decreased as curcumin and piperine concentration increased (P < 0.05). Furthermore, the result of principal component analysis indicated that curcumin and piperine markedly impeded BSA oxidation, resulting in a lower level of fAGEs in model systems. Therefore, adding curcumin and piperine may facilitate reduced fAGEs levels in BSA–fructose model.     

基于1D CNN与XGBoost的恶意代码纹理检测

恶意代码数量已经呈现爆炸式增长,对于恶意代码的检测防护显得尤为重要.近几年,基于深度学习的恶意代码检测方法开始出现,基于此,提出一种新的检测方法,将恶意代码二进制文件转化为十进制数组,并利用一维卷积神经网络(1 Dimention Convolutional Neural Networks,1D CNN)对数组进行分类和识别.针对代码家族之间数量不平衡的现象,该算法选择在分类预测上表现良好的XGBoost,并对Vision Research Lab中的25个不同恶意软件家族的9458个恶意软件样本进行了实验.实验结果表明,所提的方法分类预测精度达到了97％.