Approximations for Markovian multi-class queues with preemptive priorities

We discuss the approximation of performance measures in multi-class M/M/k queues with preemptive priorities for large problem instances (many classes and servers) using class aggregation and server reduction. We compared our approximations to exact and simulation results and found that our approach yields small-to-moderate approximation errors.     

The moduli space of flat SU (2) and SO (3) connections over surfaces

All the connected components of the moduli space of flat connections on SU (2) and SO (3) (trivial and non-trivial) bundles over closed oriented surfaces are determined. The symplectic structure and volumes of the non-maximal strata of the moduli space are also determined.     

Canonical Connections and Algebraic Ricci Solitons of Three-dimensional Lorentzian Lie Groups

In this paper, the author computes canonical connections and KobayashiNomizu connections and their curvature on three-dimensional Lorentzian Lie groups with some product structure. He defines algebraic Ricci solitons associated to canonical connections and Kobayashi-Nomizu connections. He classifies algebraic Ricci solitons associated to canonical connections and Kobayashi-Nomizu connections on three-dimensional Lorentzian Lie groups with some product structure.     

多通道反应O(3P)＋CH2F的理论研究

用密度泛函理论研究了氧原子与氟代甲基自由基的反应.反应中出现的所有物种的平衡构型用B3LYP方法在6-311＋＋G（2d, 2p）基组水平上进行了优化，同时对各物种进行了频率分析；在同一理论水平上计算了各反应通道的势能面变化，分析了反应物、中间体、过渡态、产物的振动模式随反应途径的变化关系，阐明了该多通道反应的反应机理.     

基于改进U-Net的SPECT骨显像病灶分割研究北大核心CSCD

在核医学中,单光子发射计算机断层(single-photon emission computed tomograpy,SPECT)骨显像是辅助医师诊断癌症的重要手段。针对骨显像图像信噪比低、边界模糊、病灶小难以提取和人工勾画病灶耗时等问题,提出一种基于改进U-Net网络的骨显像病灶自动分割算法。该算法在U-Net的原卷积块基础上,采用了多尺度密集连接(multi-scale dense connection,MDC)的方式来提高对小病灶特征的提取能力,同时解决了网络加深后出现的梯度消失问题。其次,为提取病灶的细节特征,在密集连接和跳跃连接处引入了注意力机制结构。最后,针对使用小样本数据集,模型难以收敛的问题,采用迁移学习的方法,优化了模型的初始参数,提升模型的泛化能力和分割效率。此外,为了降低计算量、进一步提高分割效果,对数据集进行了裁剪和去噪。同时,将处理后的图像采用旋转、镜像等方法进行了数据扩充。实验结果表明,改进的U-Net的识别精确率(precision)、平均交并比(mean intersection-over-union,mIoU)分别能达到0.7352、0.4673,效果优于目前主流的分割算法,具有一定实际应用价值。     

面向B5G/6G的三三三网络体系架构和优化学习机制

面对未来B5G/6G网络是大连接复杂智能网络,极其大量的用户连接、需求连接、服务连接,加之3G、4G、5G甚至6G的综合运用,将导致网络变得极其复杂的挑战,提出了一种三三三网络体系架构,它是一个包含3类网络(核心网、接入网、终端网)、3种资源(频率带宽、功耗、时延)和3项需求(生活、工作、服务)的三维立体综合优化的体系...     

NEAX61Σ交换机利用现有的资源提高接通率的两种方式

介绍了灵活运用NEAX61Σ交换机现有的软硬件资源 ,修改adc和rlst数据后能使用语音信箱的有关功能 ,进而提高落地接续接通率的两种方法。     

Left and right centers in quasi-Poisson geometry of moduli spaces

We introduce left central and right central functions and left and right leaves in quasi-Poisson geometry, generalizing central (or Casimir) functions and symplectic leaves from Poisson geometry. They lead to a new type of (quasi-)Poisson reduction, which is both simpler and more general than known quasi-Hamiltonian reductions. We study these notions in detail for moduli spaces of flat connections on surfaces, where the quasi-Poisson structure is given by an intersection pairing on homology.     

Multi-adjoint algebras versus non-commutative residuated structures

Adjoint triples and pairs are basic operators used in several domains, since they increase the flexibility in the framework in which they are considered. This paper introduces multi-adjoint algebras and several properties; also, we will show that an adjoint triple and its “dual” cannot be considered in the same framework.Moreover, a comparison among general algebraic structures used in different frameworks, which reduce the considered mathematical requirements, such as the implicative extended-order algebras, implicative structures, the residuated algebras given by sup-preserving aggregations and the conjunctive algebras given by semi-uninorms and u-norms, is presented. This comparison shows that multi-adjoint algebras generalize these structures in domains which require residuated implications, such as in formal concept analysis, fuzzy rough sets, fuzzy relation equations and fuzzy logic.