基于集成学习的P2P流量识别模型

将DTNB、OneR和BP神经网络算法集成用于P2P流量识别,构建了基于该集成学习算法的P2P流量识别模型。利用网络流量特征和机器学习中生成规则的集成分类算法将网络流量分为P2P流量和非P2P流量。所建立的P2P流量识别模型分为网络流量特征的获取、P2P流量特征选取以及建立流量分类模型三个步骤。采用十折交叉验证与测试集相结合的方法CTFCVWTS(combining T-fold cross validation with testing set)评估模型的合理性和提出方法的有效性。实验得出流量分类准确率平均为97.27%。结果表明,该模型具有较高的P2P流量识别准确率。

Psycho-Emotional Aspects and Risk Factors Associated with the Onerous Nature of Retirement Concerning Civil Servants within the Public Administration in Parakou (Benin)

Background: In 20% of cases, retirement is experienced as a life event with risk of isolation and loneliness. The objective is to study the psycho-emotional aspects and risk factors associated with civil servants onerous retirement. Methods: It was a descriptive and cross-sectional study which involved 233 pensioners met at the Public Treasury’s Revenue Office in Parakou from June 1 to August 30, 2013. Data were collected through an interview on vulnerability criteria, and Hamilton Depression Rating Scale was used in assessing both anxiety level and depressed mood. Results: Pensioners’ average age was 61.1 ± 6.7 years. Males aged 50 to 60 years represented 47.2%. Following the announcement of retirement, 88.9% had feelings of fear, and 75.8% showed moderate to severe anxiety prevalence of which 88.2% were females and 73.9% males. The distribution according to professional categories was as follow: A (77.3%);B (72.6%);C (69.2%);(85.5%);E (85.7%). Retirement risk factors were: dependants (73.1%), age-related disorders (40.3%), lack of health insurance coverage (24.6%) resulting in a depressed mood (36%) strongly associated with a feeling of family and social abandon (87.8% P = 0.02). Conclusion: Depending on job onerousness in each professional category and associated psycho-social risk factors, it is important to implement an adequate action plan in view of promoting pensioners mental health.

Theoretical Research on Stirling Reversible Heat Engine

A mathematical model that describes the relationship between structural parameters and operating parameters was established based on the structure of Stirling reversible heat engine by Schmidt analyses. The calculation formulas of output power and average temperature of Stirling reversible heat engine were deduced, which disclosed the relationship between average temperature and ultimate pressure. A set of theoretical calculation formula of Stirling reversible heat engine was explored by combining definition of efficiency of heat engine. Research results provide theoretical references to design of parts and control system of Stirling reversible heat eneine.

脊柱转移性肿瘤患者脊柱手术后早期相邻节段性疾病发生的分级评估

[目的]评估脊柱转移性肿瘤患者接受脊柱固定术后早期发生相邻节段疾病(adjacent segmental disease,ASD)的情况。[方法 ]回顾性分析收治的241例脊柱肿瘤患者,208例(86.3%)患者接受手术治疗,167例患者完成术后12个月的MRI和临床检查随访。根据手术方式的不同,将患者分为7组,并通过基于MRI结果的Oner分型标准和Pfirrmann分级标准诊断手术患者术后12个月内ASD发生率,并比较2种诊断标准对脊柱转移瘤术后ASD的诊断差异。[结果]根据Oner分型标准诊断ASD的发生率为11.4%,Pfirrmann分级标准诊断ASD的发生率为24.0%,两者对ASD的诊断差异有统计学意义(P<0.001),但不同手术方式患者间两种诊断标准对ASD的诊断无明显差异(P均>0.05)。基于临床症状诊断ASD的发生率为3.6%,显著低于Oner和Pfirrmann诊断标准(P=0.007、0.002)。脊柱稳定性7～9级的患者术后ASD发生率高于脊柱稳定性4～6级的患者(29.9%vs 15.7%,P=0.034)。[结论]基于影像学检查结果对脊柱转移性肿瘤患者术后早期发生ASD的诊断价值优于临床症状诊断标准,其中Pfirrmann分级标准的诊断敏感性高于Oner分型标准。