首页 | 本学科首页   官方微博 | 高级检索  
相似文献
 共查询到20条相似文献,搜索用时 109 毫秒
1.
基于慢特征重构与改进DPLS的软测量建模   总被引:1,自引:0,他引:1       下载免费PDF全文
针对过程数据中存在的噪声干扰及动态特性,提出了一种基于慢特征重构与改进DPLS的软测量建模方法。该方法首先利用慢特征分析提取变化缓慢的成分,并用于重构原始输入,同时提出一种重构相似性指标来评价重构效果,实现用尽可能少的成分刻画数据的原有趋势,减少噪声干扰;然后梳理改进DPLS方法的完整流程,并用于分析重构输入与原输出间的关系,获得的模型更符合数据间的动态关系。本方法的有效性在TE过程及脱丁烷塔过程的软测量模型中得到了验证。  相似文献   

2.
张宇  李柠  黄道 《医学教育探索》2005,(2):208-211226
对模糊C-均值聚类算法加以改进,将系统输入数据进行模糊划分,分成具有几个不同聚类中心的子集;继而引入到多模型建模过程中,针对每个子集建立相应的径向基函数(RBF)网络模型。而全局模型则由各个子模型的输出加权组合。最后通过对聚合釜反应器软测量建模的研究,表明该方法具有拟合精度高和泛化能力强的特点,验证了此多模型建模方法的有效性和快速性。  相似文献   

3.
目的 探讨在最大径>2 cm肾结石患者临床治疗中应用智能控压输尿管软镜吸引术的安全性及有效性.方法 选取45例最大径>2 cm肾结石患者作为研究对象,所有患者均运用测压吸引鞘连接平台行智能监控肾盂内压输尿管软镜吸引取石术,观察患者治疗的有效性和安全性.结果 患者的碎石时间为(29.8±14.7)min,住院时间为(4.9±1.6)d,输尿管软镜吸引术术中碎石成功率88.89%,术后4周复查时无石率为86.67%,术后8周复查时无石率为91.11%.术后共2例(4.44%)出现相关并发症,表现为发热.发热患者经过抗生素、激素及补液等对症处理后好转.结论 在最大径>2 cm的肾结石患者的临床治疗中,智能控压输尿管软镜吸引术不仅具有较高的安全性,同时还表现出了良好的有效性,无论是手术成功率,还是术后无石率,又或者是并发症率,均较理想,是一种集安全性和有效性于一体的手术方式.  相似文献   

4.
  目的  运用长程穿戴式动态心电监测记录仪, 通过分析自主神经的昼夜节律与阴阳的关系, 评价其作为一种判断人体阴阳平衡手段的可行性。  方法  选取健康受试者30例, 将长程穿戴式动态心电监测记录仪贴于受试者胸前, 同步心电监测, 连续佩戴48 h。分析48 h内心率变异性指标如窦性心搏RR间期标准差(SDNN)、RR间期标准差的平均值(SDNN index)、相邻心跳间期差值大于50 ms的个数占总心跳间期数的百分比(pNN50)、频域参数低频功率(LF)、高频功率(HF)、低频/高频比(LF/HF)等, 观察其昼夜节律变化, 分析与阴阳的关系。  结果  心率变异性各项指标呈现出昼夜变化, 反映人体的昼夜节律: 夜间SDNN、SDNN index、pNN50、HF均高于日间, 差异具有统计学意义(P < 0.05);LF/HF日间高于夜间, 差异具有统计学意义(P < 0.05)。  结论  长程穿戴式动态心电监测记录仪可以通过监测及分析心率变异性, 反映自主神经的动态变化, 从而判断人体阴阳的平衡关系。   相似文献   

5.
医用恒温箱的温度调节具有大时间常数、大惯性及时变特性,实际应用中难以用阶跃辨识法快速辨识其模型以及使用常规控制器进行有效控制。继电反馈辨识法不需要过程达到稳态即可快速辨识过程模型,自适应控制可以针对以上模型进行有效控制。提出了一种使用继电反馈辨识法进行恒温箱温度调节过程的辨识模型方法,并且利用该辨识模型设计了自适应控制器的控制器参数初值及控制律的方法;以瑞萨公司的C2B8SNFA为主控器,开发了恒温箱温度自适应调节器。实验结果验证了本文方法的可行性和有效性。  相似文献   

6.
利用Backlash描述函数的实部和虚部构造一个函数,将其进行叠加,从而建立了一个新颖的迟滞非线性模型.以所建模型的输出作为神经网络的输入信号之一,建立了一个神经网络迟滞模型.实验结果验证了所提模型的有效性.  相似文献   

7.
讨论了用寂态热力学方法处理负压系统时,将会遇到的三类难题:闭系中(火用)大于内能,(火无)为负值;稳流物系条件下,(火用)出现负值,(火无)大于焓;能级计算时可能出现正、负无限大值。指出它们来源于(火用)概念的内在矛盾及能级定义的不确切性,可以用‘物系∪环境’模型加以解决。论证了新的环境概念,指出(火用)值为零的状态唯一地对应于寂态,从而消除了负压稳流物系的所谓‘环境线’。提出了‘大系统最大输出能能级最小,最小可逆输出能能级最大’原理,根据这项原理,给出了推导‘负压物系∪环境’有关参数表达式的方法。  相似文献   

8.
分析了现有文献中寂态、(火用)及(火无)等概念。提出大系统(物系∪环境)模型。将‘寂态’重新解释为‘大系统寂态’;将大系统能量划分为‘最大可能输出能’E_M和‘不可输出能’E_N(不利用外界作功条件下);定义大系统(火用)E_X为E_M中可以转变部分,大系统(火无)为E_M中不可转变部分;E_N包括大系统可逆地过渡到寂态过程中两子系统直接互相传递的能E_i和大系统其余内能。这样,重新解释了能=(火用)+(火无)公式的意义,从而对现有文献中的一些有争议问题提出了作者的解决方法。最后推导了正压物系∪环境的E_M(物理能)E_X及A_n的表达式。  相似文献   

9.
针对间歇过程提出了基于小波神经网络的迭代学习优化控制算法,实现产品终点质量指标的控制。小波递归神经网络用于建立提供长期预测的间歇过程模型。由于模型误差以及未知干扰的影响,基于预测模型得到的控制变量在实际应用中得不到期望的终点质量指标。利用间歇过程的重复特性,采用迭代学习优化控制改善批次间的产品质量,根据以前批次的模型预测误差均值来修正神经网络模型预测输出,继而计算出下一个批次的控制输入。随着批次的进行,模型误差逐渐消失,控制输入达到最优控制。仿真实验结果验证了该算法的有效性。  相似文献   

10.
基于PUSH库存控制策略提出了在不确定生产提前期、恒定顾客需求率和产品回收率条件下的制造/再制造混合生产系统库存控制模型,可用品仓库库存由新产品制造过程和回收产品的再制造过程共同补充。不确定的生产提前期可描述为随机灰色变量,提出的随机灰色模拟技术可为不确定函数产生输入-输出数据,利用该输入-输出数据训练后的神经网络可加速不确定函数的模拟过程,由随机灰色模拟、神经网络和遗传算法集成的混合智能优化算法可求解该库存模型。数值分析结果表明:平均生产成本随给定的顾客服务水平和生产提前期的增加而增加,该不确定模型符合实际库存系统的实际情况,提出的智能优化算法可优化复杂的不确定规划问题。  相似文献   

11.
王晓辉  潘雁  朱珺 《川北医学院学报》2017,32(6):871-873,889
目的:建立老年(>65岁)肺癌患者住院化疗期间发生感染的预测模型,评价模型的预测性能,探讨个体发病风险预测的方法。方法:收集本院收治的老年肺癌化疗患者327例,按4∶1分为训练组(n=260)与检验组(n=67),分别用于筛选变量、建立预测模型及对模型的检测和评价。应用logistic回归Logistic Regression(LR)模型对资料进行多因素筛选,将筛选出有统计学意义的指标建立LR模型和人工智能神经网络(artificial neural network,ANN)模型,利用受试者工作特征曲线(receiver operator characteristic,ROC)评价模型的预测性能。结果:医院感染的发生率为21.10%(69/327),进一步分析发现骨髓抑制(≥Ⅱ度,P=0.001)、长时间住院(≥14 d,P=0.001)、糖尿病(P=0.005)、联合化疗(含顺铂,P=0.009),肺癌手术史(P=0.033)以及应用激素(P=0.025)是老年肺癌患者化疗期间发生感染的危险因素。利用两种模型对67名检验集预测,预测正确率分别为87.15%和79.81%;通过ROC曲线下面积比较模型的预测能力:LR模型曲线下面积(0.757±0.036)小于ANN模型(0.905±0.017)。结论:在老年肺癌患者住院化疗期间发生感染的预测方面ANN模型预测性能优于LR模型。  相似文献   

12.
Background  Various methods can be applied to build predictive models for the clinical data with binary outcome variable. This research aims to explore the process of constructing common predictive models, Logistic regression (LR), decision tree (DT) and multilayer perceptron (MLP), as well as focus on specific details when applying the methods mentioned above: what preconditions should be satisfied, how to set parameters of the model, how to screen variables and build accuracy models quickly and efficiently, and how to assess the generalization ability (that is, prediction performance) reliably by Monte Carlo method in the case of small sample size.
Methods  All the 274 patients (include 137 type 2 diabetes mellitus with diabetic peripheral neuropathy and 137 type 2 diabetes mellitus without diabetic peripheral neuropathy) from the Metabolic Disease Hospital in Tianjin participated in the study. There were 30 variables such as sex, age, glycosylated hemoglobin, etc. On account of small sample size, the classification and regression tree (CART) with the chi-squared automatic interaction detector tree (CHAID) were combined by means of the 100 times 57 fold stratified cross-validation to build DT. The MLP was constructed by Schwarz Bayes Criterion to choose the number of hidden layers and hidden layer units, alone with levenberg-marquardt (L-M) optimization algorithm, weight decay and preliminary training method. Subsequently, LR was applied by the best subset method with the Akaike Information Criterion (AIC) to make the best used of information and avoid overfitting. Eventually, a 10 to 100 times 3–10 fold stratified cross-validation method was used to compare the generalization ability of DT, MLP and LR in view of the areas under the receiver operating characteristic (ROC) curves (AUC).
Results  The AUC of DT, MLP and LR were 0.8863, 0.8536 and 0.8802, respectively. As the larger the AUC of a specific prediction model is, the higher diagnostic ability presents, MLP performed optimally, and then followed by LR and DT in terms of 10–100 times 2–10 fold stratified cross-validation in our study. Neural network model is a preferred option for the data. However, the best subset of multiple LR would be a better choice in view of efficiency and accuracy.
Conclusion  When dealing with data from small size sample, multiple independent variables and a dichotomous outcome variable, more strategies and statistical techniques (such as AIC criteria, L-M optimization algorithm, the best subset, etc.) should be considered to build a forecast model and some available methods (such as cross-validation, AUC, etc.) could be used for evaluation.
  相似文献   

13.
There is a conflict between Data Envelopment Analysis (DEA) theory’s requirement that inputs (outputs) be substitutable, and the ubiquitous use of nonsubstitutable inputs and outputs in DEA applications to hospitals. This paper develops efficiency indicators valid for nonsubstitutable variables. Then, using a sample of 87 community hospitals, it compares the new measures’ efficiency estimates with those of conventional DEA measures. DEA substantially overestimated the hospitals’ efficiency on the average, and reported many inefficient hospitals to be efficient. Further, it greatly overestimated the efficiency of some hospitals but only slightly overestimated the efficiency of others, thus making any comparisons among hospitals questionable. These results suggest that conventional DEA models should not be used to estimate the efficiency of hospitals unless there is empirical evidence that the inputs (outputs) are substitutable. If inputs (outputs) are not substitutes, efficiency indicators valid for nonsubstitutability should be employed, or, before applying DEA, the nonsubstitutable variables should be combined using an appropriate weighting scheme or statistical methodology.  相似文献   

14.
Data Envelopment Analysis (DEA) has been used as a performance measurement tool in efficiency assessment of healthcare systems. However, over the years, researchers and health practitioners presented the theoretical and methodological limitations of DEA that limits the full view of healthcare efficiency. To address these limitations, a commonly used strategy is to integrate other statistical methods and techniques with DEA to provide better efficiency evaluation. This paper reviews 57 studies with DEA applications in the healthcare industry to illustrate the integrated analysis of healthcare efficiency. With DEA as the central method, regression models in conjunction with statistical tests are commonly used. Input-oriented radial DEA models using predominantly capacity-related inputs and activity-related outputs and following either constant return to scale or variable return to scale assumptions are mostly applied to measure healthcare efficiency.  相似文献   

15.
年龄-时期-队列模型广泛应用于描述性流行病学来分析慢性病发病率和死亡率变化趋,它改进了传统的疾病描述性分析方法。然而模型的三因素间存在着安全线性依赖性,使参数不能得出唯一的估计值,许多学者提出了各种方法加以解决,现对近些年来国内外学者提出的年龄-时期-队列模型及其参数估计中线性依赖性问题的解决方法进行综述。  相似文献   

16.
星点设计-效应面法优化苯甲酸利扎曲普坦片的处方   总被引:9,自引:1,他引:8  
吴伟  阙俐  陈健  沈熊  夏红 《复旦学报(医学版)》2003,30(4):381-383,387
目的 星点设计-效应面优化法优化苯甲酸利扎曲普坦片的处方。方法 以微晶纤维素的用量和交联聚维酮的用量为考察因素,崩解时间为指标,用线性方程和二次及三次多项式描述崩解时间和两个影响因素之间的数学关系,根据最佳数学模型描绘效应面,选择最佳处方,并进行预测分析。结果 崩解时间与微晶纤维素的用量和交联聚维酮用量间的关系不能用线性方程描述,二次及三次多项式拟合,相关系数分别为0.9924和0.9985,具有较高的可信度。优选的最佳条件为微晶纤维索的用量为3.0g,交联聚维酮的用量为0.3g。最佳处方的崩解时间理论值与预测值差为-2.68%。结论 所建立的模型预测性良好。  相似文献   

17.
目的探讨高血压病合并急性脑梗死患者血清视黄醇结合蛋白4(RBP4)与心率变异性(HRV)之间的关系。方法选择78例高血压病合并急性脑梗死患者为观察组,以50例单纯高血压病为对照组。采用酶联免疫吸附法测定受试者血清RBP4水平,用动态心电监护仪监测HRV时域参数,包括24h所有窦性心搏R-R间期的标准差(SDNN)、24h每5分钟窦性心搏R—R问期平均值的标准差(SDANN)、24h每5分钟所有窦性心搏R-R间期标准差的均值(SDNNindex)、相邻窦性心搏R-R间期差值的均方根(rMSSD)、24h相邻窦性心搏R-R间期差值超过50ms的百分比(PNN50)。结果对照组与观察组患者血清RBP4水平分别为(15.5±5.4/24.2±7.6)mg/L,差异具有统计学意义(P<0.01)。对照组与观察组的HRV时域参数分别为SDNN(112.3±21.9/79.7±19.1)ms、SDANN(36.1±9.3/25.8±7.2)ms、SDNNindex(96.7±26.8/81.3±25.0)ms、rMSSD(23.2 4±7.9/19.4±5.7)ms、PNN50(8.9±4.1/5.7±3.2)%,观察组均低于对照组,差异具有统计学意义(P<0.05)。血清RBP4与SDNN、SDANN、SDNNindex、rMSSD、PNN50均呈显著负相关(r=-0.483、r=-0.452、r=-0.414、r=-0.293、r=-0.286,P<0.01或P<0.05)。结论血清视黄醇结合蛋白4与HRV各参数呈显著负相关,可能参与高血压病合并急性脑梗死患者自主神经功能调节失衡的形成。  相似文献   

18.
OBJECTIVE:To explore the pathologic characteristics of hyperplasia of the mammary gland(HMG) by observing differences in infrared radiation temperature of points of HMG in patients with different syndromes compared with healthy controls.METHODS:AFLIRSystems Therma CAM P30 infrared thermal camera was used to detect the infrared temperature of Shanzhong(CV 17),Qimen(LR 14),Zhongwan(CV 12),Qihai(CV 6),Guanyuan(CV 4),Taixi(KI 3),and Taichong(LR 3) in 113 patients with HMG.Of these patients,71 were placed in the Liver Qi stagnation group,34 were placed in the Dysfunction of conception and thoroughfare vessels group,and 8 were placed in the Phlegm and blood stasis in combination group.The infrared radiation temperature of each point in the patients was compared with that of healthy controls,and the differences in the infrared radiation temperatures of the points in the patients were analyzed.RESULTS:Overall,the bilateral corresponding point in both the controls and patients exhibited no significant difference in infrared radiation temperature.In all cases,the infrared radiation temperature of the points from proximal to distal tended to decrease.In a comparison of the patients and controls,the infrared radiation temperature of the trunk points Shanzhong(CV 17),Qimen(LR 14),Zhongwan(CV12),Qihai(CV6),and Guanyuan(CV 4) of the patients was higher than that of the controls,while the infrared radiation temperature of the lower extremity points Taixi(KI 3) and Taichong(LR 3) was lower than that of the controls.Of these points,Shanzhong(CV 17)(P=0.0368),Zhongwan(CV 12)(P=0.0028),Qihai(CV 6)(P=0.0085),and Guanyuan(CV4)(P=0.0018) showed significant differences.In a comparison of the corresponding point on the same side in the Liver Qi stagnation group and controls,the infrared radiation temperature of Shanzhong(CV17)(P=0.0089),right-side Qimen(LR 14)(P=0.0382),Zhongwan(CV 12)(P= 0.0000),Qihai(CV 6)(P=0.0011),and Guanyuan(CV 4)(P=0.0000) of the patients was significantly higher than that of the controls,while the differences in the infrared radiation temperature of the other points were not statistically significant(P= 0.0833-0.8397).In a comparison of the corresponding point on the same side in the Dysfunction of conception and thoroughfare vessels group and controls,the infrared radiation temperature of left-side Taichong(LR 3)(P=0.0048),right-side Taichong(LR 3)(P=0.0329),left-side Taixi(KI 3)(P= 0.0171),and right-side Taixi(KI 3)(t=0.544,P= 0.0165) of the patients was significantly lower than that of the controls,while the differences in the infrared radiation temperature of the other points were not statistically significant(P=0.3793-0.9197).In a comparison of the corresponding point on the same side in the Phlegm and blood stasis in combination group and controls,the infrared radiation temperature of Shanzhong(CV 17),Qimen(LR 14),Qihai(CV 6),Guanyuan(CV 4),Taixi(KI 3),Taichong(LR 3),and Zhongwan(CV 12) tended to increase,but without statistical significance(P=0.175-.759).CONCLUSION:The corresponding points of HMG patients with different syndromes are in different deficiency/excess states.Changes in the infrared radiation temperature of the trunk points Shanzhong(CV 17),Qimen(LR 14),Zhongwan(CV 12),Qihai(CV 6),and Guanyuan(CV 4) are closely related to the pathological characteristics of the Liver Qi stagnation syndrome of HMG patients,while changes in the infrared radiation temperature of the lower extremity points Taixi(KI 3) and Taichong(LR 3) are closely related to the pathological characteristics of the Dysfunction of conception and thoroughfare vessels syndrome of HMG patients.On the whole,HMG patients with Liver Qi stagnation syndrome are characterized by "upper excess," and those with Dysfunction of conception and thoroughfare vessels syndrome are characterized by"lowerdeficiency."  相似文献   

19.

Background

Researchers have used logistic regression (LR) and non-linear response surface models (RSMs) to predict patient responses to sedation. The reduced Greco and hierarchy RSMs have proven to be more appropriate than other RSMs in gastrointestinal endoscopies using midazolam and alfentanil. In this study, we evaluate the performance of a simpler model, LR, and compared it with that of RSM.

Methods

Thirty-three patients who received esophagogastroduodenoscopy (EGD) and colonoscopy sedation with midazolam and alfentanil were enrolled in the study. LR was performed for the EGD group and validated using the colonoscopy group. The two RSMs were performed using the same process, and performances and receiver operating characteristic (ROC) curves of the models were evaluated.

Results

The native EGD LR model had an ROC curve area of 0.94. For external validation, the ROC curves were 0.92, 0.94, and 0.94 for the reduced Greco, hierarchy, and LR models, respectively. Pairwise comparison between models was not significant.

Conclusion

The LR model performed as well as RSM in generalizing the predicted sedative effect of midazolam and alfentanil during gastrointestinal endoscopies. LR may be used for generalization across patients experiencing procedures with similar stimulus intensities.  相似文献   

20.
目的 应用深度学习模型循环神经网络(recurrent neural network,RNN)及其变体门控循环单元(gated recurrent unit,GRU),基于临床真实数据,构建腹膜透析临床预后预测模型,并比较其与医学研究中常用的逻辑回归(logistic regression, LR)模型的预测性能,探索预测结果中可能的医学意义。方法 使用北京大学第三医院腹膜透析门诊的常规诊疗数据,基于患者在开始透析时的基线数据、随访数据和预后数据构建RNN和GRU预测模型。使用受试者工作特征曲线下面积(area under the ROC curve,AUROC)、召回率(recall)、F1分数(F1-score)三个指标在测试集上评价比较模型对患者死亡风险的预测效果。结果 共纳入656例患者,其中死亡患者261例,共计13 091条诊断记录。经过十折交叉验证调整超参数并在单独的测试集测试结果显示,LR模型、RNN模型、GRU模型的AUROC分别为0.701 4、0.786 0、0.814 7,RNN和GRU模型的预测性能显著优于传统的LR模型。在召回率和F1分数方面,RNN和GRU模型的性能也均显著优于LR模型,且GRU模型表现最好。进一步分析显示GRU模型在不同预测窗口下对于不同死因或相同死因的召回率不尽相同。结论 RNN模型(尤其是GRU模型)相比于传统医学研究所使用的LR模型,对于腹膜透析临床预后预测具有更佳效果,可能有助于医生早期干预,提高医疗质量,具有很强的临床应用价值。  相似文献   

设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号