Randomized Latent Factor Model for High-dimensional and Sparse Matrices from Industrial Applications

Latent factor(LF) models are highly effective in extracting useful knowledge from High-Dimensional and Sparse(HiDS) matrices which are commonly seen in various industrial applications. An LF model usually adopts iterative optimizers,which may consume many iterations to achieve a local optima,resulting in considerable time cost. Hence, determining how to accelerate the training process for LF models has become a significant issue. To address this, this work proposes a randomized latent factor(RLF) model. It incorporates the principle of randomized learning techniques from neural networks into the LF analysis of HiDS matrices, thereby greatly alleviating computational burden. It also extends a standard learning process for randomized neural networks in context of LF analysis to make the resulting model represent an HiDS matrix correctly.Experimental results on three HiDS matrices from industrial applications demonstrate that compared with state-of-the-art LF models, RLF is able to achieve significantly higher computational efficiency and comparable prediction accuracy for missing data.I provides an important alternative approach to LF analysis of HiDS matrices, which is especially desired for industrial applications demanding highly efficient models.

Robust Latent Factor Analysis for Precise Representation of High-Dimensional and Sparse Data

High-dimensional and sparse(HiDS)matrices commonly arise in various industrial applications,e.g.,recommender systems(RSs),social networks,and wireless sensor networks.Since they contain rich information,how to accurately represent them is of great significance.A latent factor(LF)model is one of the most popular and successful ways to address this issue.Current LF models mostly adopt L2-norm-oriented Loss to represent an HiDS matrix,i.e.,they sum the errors between observed data and predicted ones with L2-norm.Yet L2-norm is sensitive to outlier data.Unfortunately,outlier data usually exist in such matrices.For example,an HiDS matrix from RSs commonly contains many outlier ratings due to some heedless/malicious users.To address this issue,this work proposes a smooth L1-norm-oriented latent factor(SL-LF)model.Its main idea is to adopt smooth L1-norm rather than L2-norm to form its Loss,making it have both strong robustness and high accuracy in predicting the missing data of an HiDS matrix.Experimental results on eight HiDS matrices generated by industrial applications verify that the proposed SL-LF model not only is robust to the outlier data but also has significantly higher prediction accuracy than state-of-the-art models when they are used to predict the missing data of HiDS matrices.

Human Factor Based Leadership: Critical Leadership Tools to Reduce Burnout and Latent Error in a Time of Accelerating Change

The majority of errors in healthcare are from systems factors that create the latent conditions for error to occur. The majority of occupational stressors causing burnout are also the result of systemic factors. Advances in technology create new levels of stress and expectations on healthcare workers (HCW) with an endless infusion of requirements from multiple authoritative sources that are tracked and monitored. The quality of care and safety of patients is affected by the wellbeing of HCWs who now practice in an environment that has become more complex to navigate, often expending limited neural resource (brainpower) on classifying, organizing, constantly making decisions on how and when they can accomplish what is required(extraneous cognitive load) in addition to direct patient care. New information demonstrates profound biological impact on the brains of those who have burnout in areas that affect the quality and safety of the decisions they make-which affects risk to patients in healthcare. Healthcare administration curriculum currently does not include ways to address these stress-induced problems in healthcare delivery. The science of human factors and ergonomics (HFE) promotes system performance and worker wellbeing. Patient safety is one component of system performance. Since many requirements come without resource to accomplish them, it becomes incumbent upon health system leadership to organize the means for completion of these to minimize the needless loss of brain power diverted away from the delivery of patient care. Human Factor-Based Leadership (HFBL) is an interactive, problem solving seminar series designed for healthcare leaders. The purpose is to provide relevant human factor science to integrate into their leadership and management decisions to make HCWs occupational environment more manageable and sustainable-which makes safer conditions for clinician wellbeing and patient care. After learning the content, a cohort of healthcare leaders believed that adequately addressing HFE in healthcare delivery would significantly reduce clinician burnout and risk of latent errors from upstream leadership decisions. An overview of the content of the seminars is described. Leadership feedback on usability of these seminars is reported. Three HFBL seminars described are Human Factor Relevance in Leadership, Biopsychosocial Approach to Wellness and Burnout, Human Factor Based Leadership: Examples and Applications.

Nuclear factor κB represses the expression of latent membrane protein 1 in Epstein-Barr virus transformed cells

AIM: To investigate the role of nuclear factor κB(NF-κB) in the regulation of Epstein-Barr virus(EBV) latent membrane protein 1(LMP1) in EBV transformed cells. METHODS: LMP1 expression was examined in EBV transformed human B lymphocytes with modulation of NF-κB activity. RESULTS: EBV infection is associated with several human cancers. EBV LMP1 is required for efficient transformation of adult primary B cells in vitro, and is expressed in several pathogenic stages of EBVassociated cancers. Regulation of EBV LMP1 involves both viral and cellular factors. LMP1 activates NF-κB signaling pathway that is a part of the EBV transformation program. However, the relation between NF-κB and LMP1 expression is not well established yet. In this report, we found that blocking the NF-κB activity by Inhibitor of κB stimulated LMP1 expression, while the overexpression of NF-κB repressed LMP1 expression in EBV-transformed IB4 cells. In addition, LMP1 repressed its own promoter activities in reporter assays, and the repression was associated with the activation of NF-κB. Moreover, NF-κB alone is sufficient to repress LMP1 promoter activities. CONCLUSION: Our data suggest LMP1 may repress its own expression through NF-κB in EBV transformed cells and shed a light on LMP1 regulation during EBV transformation.

Latent tuberculosis:Risk factors,screening and treatment in liver transplantation recipients from an endemic area

BACKGROUND Patients undergoing solid organ transplantation,particularly those who live or have lived in tuberculosis(TB)endemic areas,are at a high risk of developing TB.The majority of post-transplantation TB cases are associated with reactivation of latent TB infection(LTBI).Brazil is in a single position with overlapping areas of high TB endemicity and high transplant activity.In liver transplant(LT),one should be aware of the potential hepatotoxicity associated with the treatment regimens for LTBI.AIM To evaluate the frequency of LTBI in LT patients and treatment-related issues.METHODS This was a retrospective analysis of a cohort of cirrhotic patients aged≥18 years,who underwent LT at a high-complexity teaching hospital from January 2005 to December 2012.RESULTS Overall,429 patients underwent LT during the study period.Of these,213(49.7%)underwent the tuberculin skin test(TST)during the pre-transplant period,and 35(16.4%)of them had a positive result.The treatment for LTBI was initiated after LT in 12(34.3%)of the TST-positive patients;in 3(25.0%),treatment was maintained for at least 6 mo.CONCLUSION The prevalence of LTBI was lower than expected.Initiation and completion of LTBI treatment was limited by difficulties in the management of these special patients.

Latent Tuberculosis Infection among Household Contacts of Pulmonary Tuberculosis Cases in Central State, Sudan: Prevalence and Associated Factors

Introduction: Tuberculosis is a major health problem in developing countries including Sudan. Screening for TB cases through Household contacts (HHCs) investigation is an appropriate strategy to interrupt transmission of TB. Objectives: To determine the prevalence tuberculosis infection and risk factors for tuberculosis infection among household contacts in Wadimadani locality, Central State, Sudan, between November 2015 and April 2016. Methods: An analytical cross-sectional study conducted. During study period, to confirm TB diagnosis, all suspect contacts were tested through sputum samples, tuberculin skin test or chest X-ray. Structured questionnaire was used to collect socio-demographic and environmental factors. Results: One hundred forty six patients of smear-positive pulmonary tuberculosis were included in the study, 657 household contacts were identified and screened. Forty three new TB cases were detected from household contacts, yielding a prevalence of 6.5% (95% confidence interval = 0.05, 0.09) of latent tuberculosis infection (LTBI). Two factors were significantly associated with LTBI among HHCs: duration of contact with a TB patient ≤ 4 months (P = 0.03) and the educational status (P = 0.02). Conclusion: Screening of HHCs of index case of TB will contribute in early detection and treatment of new cases, and considered as a forward step towards eliminating TB.

Network autoregression model with grouped factor structures

Network autoregression and factor model are effective methods for modeling network time series data.In this study,we propose a network autoregression model with a factor structure that incorporates a latent group structure to address nodal heterogeneity within the network.An iterative algorithm is employed to minimize a least-squares objective function,allowing for simultaneous estimation of both the parameters and the group structure.To determine the unknown number of groups and factors,a PIC criterion is introduced.Additionally,statistical inference of the estimated parameters is presented.To assess the validity of the proposed estimation and inference procedures,we conduct extensive numerical studies.We also demonstrate the utility of our model using a stock dataset obtained from the Chinese A-Share stock market.

Collaboration Filtering Recommendation Algorithm Based on the Latent Factor Model and Improved Spectral Clustering

Due to the development of E-Commerce, collaboration filtering (CF) recommendation algorithm becomes popular in recent years. It has some limitations such as cold start, data sparseness and low operation efficiency. In this paper, a CF recommendation algorithm is propose based on the latent factor model and improved spectral clustering (CFRALFMISC) to improve the forecasting precision. The latent factor model was firstly adopted to predict the missing score. Then, the cluster validity index was used to determine the number of clusters. Finally, the spectral clustering was improved by using the FCM algorithm to replace the K-means in the spectral clustering. The simulation results show that CFRALFMISC can effectively improve the recommendation precision compared with other algorithms.

基于加速无约束张量隐因子分解模型的Web服务Qo S估计

针对基于张量非负隐因子分解模型的Web服务QoS估计方法过于依赖非负初始随机数据以及特意设计的非负训练方法,导致模型的兼容性和扩展性不高的问题,提出了加速无约束张量隐因子分解模型。其主要思想包括三部分:将非负性约束从决策参数转移到输出的隐因子,并通过单元素映射函数连接它们;运用结合动量方法的随机梯度下降算法,有效提高模型的收敛速度与估计精度;给出加速无约束张量隐因子分解模型的详细算法和结果分析。在实际工业应用中的2个动态QoS数据集上的实证研究表明,与最先进的QoS估计模型相比,所提模型具有较高的计算效率和估计精度。

山东省高校临床实习护生工作准备度的潜在剖面分析及其影响因素

目的:探讨山东省高校实习护生工作准备度现状,分析不同类别护生工作准备度的潜在特征差异,并分析其影响因素,为制定针对性的工作准备度干预方案提供参考。方法:2023年9月—10月,采用一般人口学资料调查表、工作准备度量表、反馈寻求行为量表对山东省不同地区1 478名实习护生进行调查,并对工作准备度进行潜在类别分析。结果:1 478名实习护生工作准备度可分为低准备度组(26.05%)、中准备度组(41.88%)、高准备度组(32.07%)3个类别。与高准备度组比较,女性、反馈寻求行为越低,越容易进入高准备度组;与高准备度组比较,女性、本科学历、反馈寻求行为越低,越容易进入中准备度组;与中准备度组相比,寻求反馈行为越少越容易进入低准备度组。结论:山东省实习护生工作准备度水平可分为3类,护理管理者可根据不同类型实习护生的临床特征,提高实习护生反馈寻求行为,从而提升实习护生的工作准备度。

血清LTBP2、TM4SF1在大肠癌患者预后评估中的研究

目的探讨血清转化生长因子结合蛋白2(LTBP2)、四次跨膜蛋白1(TM4SF1)在大肠癌患者预后评估中的意义。方法将南阳市中心医院2016年7月至2019年7月手术治疗的大肠癌患者108例作为试验组,同期108例健康体检者作为对照组。采用酶联免疫吸附试验(ELISA)检测两组血清LTBP2、TM4SF1水平;比较不同血清LTBP2和TM4SF1表达水平的大肠癌患者临床资料的差异;采用免疫组化染色法分析大肠癌组织中LTBP2、TM4SF1的表达;Pearson相关分析明确血清LTBP2和TM4SF1表达的相关性;受试者工作特征(ROC)曲线分析血清LTBP2和TM4SF1联合评估大肠癌患者预后的价值。结果试验组血清LTBP2和TM4SF1水平均高于对照组(P<0.05),二者表达水平呈正相关(r=0.305,P<0.05),大肠癌组织LTBP2和TM4SF1表达阳性率高于癌旁组织(P<0.05)。肿瘤直径>5cm、肿瘤低分化、有淋巴结转移、TNM分期Ⅲ~Ⅳ期、有脉管瘤栓的大肠癌患者血清LTBP2和TM4SF1表达水平高于患者肿瘤直径≤5 cm、肿瘤中高分化、无淋巴结转移、TNM分期Ⅰ~Ⅱ期、无脉管瘤栓的大肠癌患者(P<0.05)。血清LTBP2、TM4SF1联合预测大肠癌患者预后不良的AUC为0.886。结论大肠癌患者血清LTBP2和TM4SF1水平升高,二者联合对大肠癌患者预后不良具有较好的预测价值。

缺血性脑卒中病人心理韧性的潜在剖面分析

目的:探讨缺血性脑卒中病人心理韧性的潜在剖面特征,并分析不同剖面病人的特征差异。方法:采用便利抽样法于2023年3月—7月选取天津市某三级甲等神经专科医院门诊就诊的缺血性脑卒中病人作为调查对象。采用一般资料调查问卷、心理韧性量表、脑卒中病人自我效能量表和领悟社会支持量表进行问卷调查。对调查对象的心理韧性特征进行潜在剖面分析,并采用单因素分析和有序多分类Logistic回归分析不同剖面的影响因素。结果:233例缺血性脑卒中病人的心理韧性特征分为3个潜在剖面,分别是低水平心理韧性组(17.60%)、中等水平心理韧性组(36.91%)和高水平心理韧性组(45.49%)。有序多分类Logistic回归分析结果显示,自我效能(OR=1.094,P<0.001)、领悟社会支持(OR=1.130,P<0.001)、小学及以下学历(OR=0.399,P=0.039)和存在多重用药(OR=0.358,P=0.037)是心理韧性潜在剖面的影响因素。结论:缺血性脑卒中病人心理韧性有明显的分类特征,临床工作者可以根据不同类别的人群特征和影响因素,给予个体化干预,以提升病人的心理韧性水平。

鉴定LTBP1作为胃癌预后的生物标志物及其与肿瘤免疫微环境的相关性

目的探索潜在转化生长因子β结合蛋白1(latent transforming growth factor beta binding protein 1,LTBP1)在胃癌发生发展及免疫微环境中的生物学功能。方法在TCGA数据库中分析LTBP1在泛癌中的表达,然后通过胃癌组织和癌旁组织进行验证。通过回归比例分析法分析LTBP1表达与临床病理变量之间的相关性。Cox回归分析和Kaplan-Meier图用于评估LTBP1在胃癌中的预后价值。通过TIMER分析LTBP1的表达水平与胃癌中免疫细胞浸润之间的相关性。免疫组织化学染色检测LTBP1蛋白在胃癌组织及癌旁组织中的表达水平。结果与正常胃组织相比,胃癌组织中LTBP1表达显著上调。其表达与病理TNM分期显著相关。LTBP1高表达的胃癌患者的总体生存率(overall survival,OS)缩短。免疫组化结果显示,与癌旁组织相比,LTBP1在胃癌组织中显著高表达。TIMER检测发现LTBP1的表达与3种免疫细胞浸润呈正相关。结论LTBP1可能是胃癌预后的一个潜在生物标志物,并影响癌症的肿瘤免疫微环境。

危险货物运输驾驶人风险倾向分类及识别模型研究

为合理评估危险货物运输驾驶人驾驶过程中的风险倾向,建立危险货物运输驾驶人风险倾向聚类及辨识体系,以动态监控系统中记录的驾驶人实时违规预警数据为基础,选取可能引发交通冲突的安全关键事件为特征参数,利用探索性因子分析方法实现指标降维,提取驾驶人风险倾向主因子,并通过K means算法聚类不同风险倾向的驾驶人,最后基于聚类结果监督训练随机森林模型,辨识未知驾驶人的风险倾向。结果表明,利用选取的8类安全关键事件特征参数,可以将驾驶人风险倾向划分为攻击驾驶倾向、鲁莽驾驶倾向、驾驶分神倾向和驾驶疲劳倾向,且可以识别风险较低的驾驶人,基于随机森林模型的驾驶人风险倾向识别准确率为88.68%,可以较好地实现危险货物运输驾驶人风险倾向辨识。研究结果为危险货物运输驾驶人风险倾向分类及识别提供了方法依据。

中青年脑卒中患者重返工作自我效能潜在剖面分析及影响因素研究

目的探讨中青年脑卒中患者重返工作自我效能(return-to-work self-efficacy,RTW-SE)的潜在分型并分析不同类别中青年脑卒中患者的特征差异。方法2020年11月至2021年4月,选取郑州市两个社区中共350例中青年脑卒中患者为研究对象,采用一般资料调查表、改良版Rankin量表、RTW-SE量表、疲劳严重程度量表、患者健康问卷-9、广泛性焦虑量表对其进行调查。采用潜在剖面分析患者RTW-SE的潜在分型,采用多元Logistic回归分析评估各种因素对不同分型的影响。结果中青年脑卒中患者RTW-SE分为3个类别:高内在应对-低工作应对组(19.9%);高工作应对-低内在应对组(55.8%);广泛中等水平组(24.3%)。婚姻状况、家庭人均月收入、职业类型、患病前是否为家庭主要经济支持、卒中后疲劳程度、功能障碍是影响中青年脑卒中患者RTW-SE潜在剖面分类的预测因素(均P<0.05)。结论中青年脑卒中患者RTW-SE具有3个潜在类别,医护人员应注意不同患者的RTW-SE特征,实施有针对性的干预措施,帮助其尽快重返社会。

社区高血压患者收缩压波动轨迹及其影响因素的研究

目的分析和构建社区高血压人群收缩压(systolic blood pressure,SBP)变化的轨迹模型,并分析不同SBP轨迹的影响因素。方法本研究基于社区回顾性队列,运用潜类别轨迹模型(latent class trajectory modelling,LCTM)分析社区高血压人群SBP的变化模式,识别、构建SBP的纵向变化轨迹;运用无序多分类logistic回归分析不同SBP轨迹的影响因素,根据先验知识使用“有向无环图”识别和调整不同的混杂因素。结果共793名高血压患者被纳入分析,LCTM拟合的社区高血压患者SBP轨迹最优分组为3组,分别为低水平平稳组(n=561,70.74%)、下降组(n=170,21.44%)和上升组(n=62,7.82%);年龄、锻炼频率、随访方式、摄盐情况、遵医行为、有无转诊在不同SBP轨迹亚组中分布存在统计学差异(P<0.05);无序多分类logistic回归分析结果显示,以低水平平稳组为对照,“男性”、“门诊随访”的患者被分类到下降组的可能性较高,OR及95%CI分别为1.436(1.016~2.030)、1.702(1.202~2.410);而“年龄≥65岁”,“不锻炼或偶尔锻炼”,摄盐情况为“中”和“重”度的人群,被分类到上升组的可能性更高,OR及95%CI依次为1.949(1.145~3.317)、2.284(1.305~3.998)、2.433(1.272~4.654)、4.540(1.291~15.963)。结论社区高血压人群收缩压变化轨迹可分为3组,即“低水平平稳组”、“下降组”和“上升组”;性别、年龄、摄盐情况、锻炼频率、随访方式可能是收缩压轨迹的影响因素。

湖南省儿科护士信息胜任力现状及潜在剖面分析

目的:探讨湖南省儿科护士信息胜任力现状,并分析其潜在剖面和影响因素。方法:2023年6月—9月采用便利抽样法选取湖南省湘东、湘西、湘南、湘北地区各1所共4所三级甲等医院的458名儿科护士作为研究对象,采用一般资料调查表、护士信息胜任力评估量表、护士使用移动护理信息系统满意度调查表、临床护理信息系统有效性评价量表进行调查。对儿科护士信息胜任力进行潜在剖面分析,并采用Logistic回归分析识别其潜在剖面的影响因素。结果:湖南省儿科护士信息胜任力得分为(71.83±18.45)分,其信息胜任力分为3个剖面,分别为低信息胜任力组(31.4%)、中低信息胜任力组(57.2%)、高信息胜任力组(11.4%)。Logistic回归分析显示,年龄、学历、工作年限、职务、职称、是否值夜班、移动护理信息系统满意度和临床护理信息系统有效性是儿科护士信息胜任力潜在剖面类别的影响因素(P<0.05)。结论:湖南省儿科护士信息胜任力有待提升,同时存在异质性。学校和医院需重视信息胜任力在护士职业生涯中的重要性,并根据不同潜在剖面的影响因素对儿科护士进行个体化干预。

慢性心力衰竭患者衰弱变化轨迹及其影响因素

目的探索慢性心力衰竭患者衰弱变化轨迹及其影响因素,为针对性干预提供参考。方法采用方便抽样法选取住院治疗的105例慢性心力衰竭患者,对其入院至出院后6个月期间分为4个时间点进行衰弱评估,了解衰弱变化轨迹,分析影响因素。结果4个时间点患者衰弱评分分别为2(1,3)分、2(0.3)分、1(0,3)分、1(0,3)分;衰弱轨迹分为低水平缓降组(47.6%)、中水平下降组(18.1%)、中水平上升组(34.3%)3个潜类别;多分类logistic回归分析结果显示,年龄、身体质量指数、共病种数、血红蛋白、心功能分级、抑郁和体力活动是患者衰弱轨迹的影响因素(均P<0.05)。结论慢性心力衰竭患者经住院治疗护理,衰弱水平总体呈好转趋势,衰弱轨迹分为3个潜类别,轨迹具有异质性;护理人员可针对不同类别、不同阶段及影响因素制定干预计划,实现精准护理。

基于“伏毒”学说审因论治结直肠癌

伏毒作为介导结直肠癌发病、转移和复发的重要病理因素,贯穿结直肠癌发生发展的多个阶段,根据伏毒成因不同,可分为外淫伏毒和内生伏毒,审因论治结直肠癌具有重要临床意义。祛除外淫伏毒,以宣透疏散为要;对于内生伏毒,兼顾扶正与祛邪。扶正者,须注重健脾益肾、气血双补、通补兼施;祛邪者,当针对不同伏毒,分别予以燥运除伏湿、健脾化伏痰、活血化伏瘀等法。运用“伏毒”学说可对结直肠癌进行辨证求因、审因论治,以把握时机,标本兼治。

腹膜透析患者容量管理行为的异质性及影响因素分析

目的探索腹膜透析患者容量管理行为的群体异质性,并分析其人群特征和影响因素,为实施精准护理干预提供参考。方法以便利抽样法招募266例腹膜透析患者,采用一般资料调查表、持续非卧床腹膜透析患者容量管理行为量表、医院焦虑抑郁量表、家庭关怀度指数问卷进行调查。运用潜在剖面分析进行异质性检验,探索腹膜透析患者容量管理行为的潜在分类,采用无序多分类logistic回归分析进一步探讨潜在类别的影响因素。结果腹膜透析患者容量管理行为存在3种潜在类别,即容量管理消极型(15.8%)、容量管理普通型(46.6%)、容量管理积极型(37.6%)。无序多分类logistic回归分析结果显示,文化程度偏低、透析频率>3袋/d的患者归属于容量管理普通型的概率更高;居住在城市、无焦虑抑郁症状、家庭功能良好的患者更倾向于发展为容量管理积极型(均P<0.05)。结论腹膜透析患者容量管理行为存在明显的群体异质性,医护人员可根据人口学和疾病特征识别不同群体患者,并针对各群体特征和影响因素进行精准护理干预,改善居家腹膜透析患者容量超负荷,提高透析效果。