基于Bi-LSTM的多层面隐喻识别方法

以双向长短期记忆网络(Bi-LSTM)为核心,结合多层卷积神经网络以及单向长短期记忆网络构建了多层面隐喻识别模型.基于多特征协同作用的思想,利用依存关系特征、语义特征、词性特征等多特征融合输入方法,丰富了模型的学习信息.为降低信息干扰,利用基于统计学的规范化文本输入方法提升模型识别效果.在英文语料词层面和句层面实验中,各个特征均表现出明显的正向作用.裁剪和填充处理及多特征协调作用在英文语料词层面研究中使F₁值分别提升2.5%和5.1%,在句层面研究中F₁值分别提升3.1%和1.9%.在中文语料句层面实验中,最优效果的F₁值可达88.8%.     

一类高维半线性热传导方程Cauchy问题解的唯一性与稳定任

文献〔1〕用引进积分的方法讨论了一类高维半线性热传导方程混合问题解的唯一性与稳定性.本文继续文献〔1〕仍用引进积分的方法推导出同类方程Cauchy问题解的唯一性与稳定性.     

Why eliminativism?

A central thesis of Steven French's brand of ontic structural realism has always been his eliminativism about objects. Unsurprisingly, this bold and controversial thesis has seen a lot of critical discussion. In his book The Structure of the World—Metaphysics & Representation, French accordingly defends this thesis against a range of challenges. A novel feature of this defense is the use of dependence relations to articulate his eliminativism. In this paper I take a critical look at French's defense of eliminativism and argue that the dependence relations invoked do not eliminate objects.     

The tragedy of the canon; or,path dependence in the history and philosophy of science

We have previously argued that historical cases must be rendered canonical before they can plausibly serve as evidence for philosophical claims, where canonicity is established through a process of negotiation among historians and philosophers of science (Bolinska and Martin, 2020). Here, we extend this proposal by exploring how that negotiation might take place in practice. The working stock of historical examples that philosophers tend to employ has long been established informally, and, as a result, somewhat haphazardly. The composition of the historical canon of philosophy of science is therefore path dependent, and cases often become stock examples for reasons tangential to their appropriateness for the purposes at hand. We show how the lack of rigor around the canonization of case studies has muddied the waters in selected philosophical debates. This, in turn, lays the groundwork for proposing ways in which they can be improved.     

ARMA序列的样本均值和自相关函数的分布的一致收敛速度

设{x_t}是ARMA序列,其谱密度函数为g(w),自相关函数为r_k,且记A=sum from i=-∞ to +∞(r_i~2+r_(i-k)r_(i+k));又x_1,x_2,…,x_N是来自{x_t}的一段样本,样本均值和自相关函数分别是(?)和(?)_k,记N~(1/2)(2πg(0))~(-1)x和(N-R)~(1/2)A~(-1/2)((?)_k-r_k)的分布函数分别为F_N(x)和G_N(x),在一定条件下我们证明了(?)|F_N(x)-Φ(x)|≤C_1N~(-1/2),(?)|G_N(x)-Φ(x)|≤C_2(lnN)~2N~(-1/5)。其中C_1,C_2,均为常数,Φ(x)为标准正态分布,这对评估统计推断的精确度具有一定的作用。     

相关结构下的两重复合状态生命模型

研究在同单调相依结构下两重复合状态生命模型,比较了具有已知边际分布的二元分布相关序,得到了此相关结构下复合状态生命模型以及相应的剩余寿命随机变量的随机界.     

一种Fortran程序自动向量化的方法

本文从研究数据相关性出发,引进了适用于工程实现的语句间时序相关图。给出了定义,阐述了建立和化简该图的方法。本图准确地描述了语句间的时序相关性,因而可以方便地确定可向量化的语句和执行顺序,对两维和多维相关进行了分析,提出了相关判定公式,本文为自动向量化Fortran程序提供了一种理论依据和方法。     

索赔盈余风险模型中精确大偏差

考虑了控制变化族(D族)上索赔过程与保费过程构成的索赔盈余风险模型,研究了此风险模型中带相依关系的随机变量的非随机和与随机和的尾概率渐近问题,利用求相依不同分布的随机变量的非随机和与随机和的精确大偏差方法,得到了带上延拓负相依和混合相依关系的不同分布的随机变量构成的索赔风险模型中的非随机和与随机和的精确大偏差渐近的结论,最后建立了索赔盈余风险模型中精确大偏差的渐近公式.     

紧固件通孔及沉孔尺寸的简易设计

常用紧固件通孔及沉孔的尺寸，与装配的螺栓或螺钉的公称直径均有显著的线性相关关系，从而建立起各种一元线性回归方程，以取代传统的查手册取值法；若用在CAD绘图软件的运行中，更能提高绘图设计的工效。     

多孔介质中Brinkman-Forchheimer模型的结构稳定性

利用能量不等式的方法, 并借助一些先验估计, 给出多孔介质中溶解度与温度有关Brinkman-Forchheimer方程组的解对边界系数的连续依赖性和收敛性结果. 结果表明, 该类方程组对边界系数具有结构稳定性.