Quantitative headspace analysis of volatiles emitted by plants or any other living organisms in chemical ecology studies generates large multidimensional data that require extensive mining and refining to extract usef...Quantitative headspace analysis of volatiles emitted by plants or any other living organisms in chemical ecology studies generates large multidimensional data that require extensive mining and refining to extract useful information. More often the number of variables and the quantified volatile compounds exceed the number of observations or samples and hence many traditional statistical analysis methods become inefficient. Here, we employed machine learning algorithm, random forest (RF) in combination with distance-based procedure, similarity percentage (SIMPER) as preprocessing steps to reduce the data dimensionality in the chemical profiles of volatiles from three African nightshade plant species before subjecting the data to non-metric multidimensional scaling (NMDS). In addition, non-parametric methods namely permutational multivariate analysis of variance (PERMANOVA) and analysis of similarities (ANOSIM) were applied to test hypothesis of differences among the African nightshade species based on the volatiles profiles and ascertain the patterns revealed by NMDS plots. Our results revealed that there were significant differences among the African nightshade species when the data’s dimension was reduced using RF variable importance and SIMPER, as also supported by NMDS plots that showed S. scabrum being separated from S. villosum and S. sarrachoides based on the reduced data variables. The novelty of our work is on the merits of using data reduction techniques to successfully reveal differences in groups which could have otherwise not been the case if the analysis were performed on the entire original data matrix characterized by small samples. The R code used in the analysis has been shared herein for interested researchers to customise it for their own data of similar nature.展开更多
BACKGROUND: Convalescence is an important stage of stroke treatment. A lot of patients have somatic and mental disorders at various degrees. The primary standard can only reflect partial conditions of somatic disorder...BACKGROUND: Convalescence is an important stage of stroke treatment. A lot of patients have somatic and mental disorders at various degrees. The primary standard can only reflect partial conditions of somatic disorder; in addition, multiple dimensions of patients at the phase of stroke convalescence are further observed by using a lot of standards, such as signs and symptoms of traditional Chinese medicine, daily activity and psychological status. OBJECTIVE: To analyze the outcome assessments of the cases of stroke convalescence measured with different criteria consisting of various dimensions by a cross-sectional investigation of the condition of stroke convalescent patients. DESIGN: Scale evaluation. SETTING: Departments of Clinical Epidemiology Exploratory Development and Neurology, the Second Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine; National Center for Training of Design, Measurement and Evaluation in Clinical Research,Guangzhou University of Traditional Chinese Medicine. PARTICIPANTS: A total of 194 stroke convalescent patients treated in the Second Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine from July 26, 2000 to February 28, 2001 were taken as subjects of the study. There were 126 males and 68 females aged from 40 to 89 years, and the illness course ranged from 14 to 181 days. All patients met diagnosis-treatment criteria of stroke (the second version)[DTCS(V2.0)] and various kinds of diagnostic points of cerebrovascular diseases; moreover, all patients provided confirmed consents. METHODS: They were assessed by assessment methods including the following assessment instruments: DTCS(V2.0), self-designed scale of traditional Chinese medicine (TCM) symptoms (28 symptoms and physical signs were scored as 0, 1, 2 marks from none to severity), modified Edinburgh-Scandinavia stroke scale (a total of 45 marks, 0 to 15 marks as mild defect, 16 to 30 as moderate defect, 31 to 45 as severe defect), modified Barthel activities of daily life (ADL) index (a total of 100 marks, less than 60 marks as unable self-care), vitality and mental health (subscales derived from Health Survey Questionnaire, SF-36). The collected data from scales and inter-scale correlation were processed by the statistic methods mainly including descriptive analysis, Spearmen correlation analysis, factor analysis, etc. MAIN OUTCOME MEASURES: ① Average scores of scales and criteria; ② correlation between modified Edinburgh-Scandinavia stroke scale and other scales. RESULTS: All of the patients completed the assessment, and analyzed in the final analysis. ① The average scores of the scales and criteria: The average scores of DTCS(V2.0), self-designed scale of TCM symptoms, modified Edinburgh-Scandinavia stroke scale, modified Barthel ADL index, vitality and mental health scales were 6.51±6.29, 13.73±6.97, 7.56±7.35, 63.58±23.68, 52.79±23.32 and 62.83±22.75 respectively. ② Correlation between modified Edinburgh-Scandinavia stroke scale and other scales: The Spearman correlation coefficients (R ’) of modified Edinburgh-Scandinavia stroke scale with diagnosis-treatment criteria of stroke, scales of TCM symptoms, modified Barthel ADL index, vitality scale and mental health scale were 20.885, 0.302, -0.824, -0.294 and -0.258 respectively. CONCLUSION: The modified Edinburgh-Scandinavia stroke scale and DTCS(V2.0) shared the same assessment dimension, so they can be mutually alternated in some clinical practices. Discrepancy in measurements of health status was gained due to the diverse dimensions applied in outcome assessments. It is necessary to build up a multi-dimensional assessment criteria system, such as signs and symptoms, daily activities and psychological status, for assessing the stroke convalescent cases in a more comprehensive scope and reflecting the efficacy of TCM treatment scientifically.展开更多
针对液晶显示器(LCD)面板的“Chip/FPC on Glass”(C/FOG)工艺生产制造过程中存在的计量延迟大、生产异常无法提前预测的问题,本文提出一种基于神经网络的C/FOG工艺生产制造虚拟计量方法。该方法利用生产机台上的传感器采集生产过程中...针对液晶显示器(LCD)面板的“Chip/FPC on Glass”(C/FOG)工艺生产制造过程中存在的计量延迟大、生产异常无法提前预测的问题,本文提出一种基于神经网络的C/FOG工艺生产制造虚拟计量方法。该方法利用生产机台上的传感器采集生产过程中的过程状态数据,构建基于多尺度一维卷积及通道注意力模型(MS1DC-CA)的虚拟计量模型。通过多个尺度的卷积核提取不同尺度范围内的状态数据特征。在对含有缺失值的原始数据预处理中,提出了基于粒子群算法改进的K近邻填补方法(PSO-KNN Imputation)进行缺失值填充,保留特征的同时,减少因填充值引入的干扰。最后在实际生产采集的数据上进行实验对比分析,实际不良率主要集中在0.1%~0.5%,该虚拟计量模型的拟合均方误差为0.397 7‱,低于其他现有拟合模型,在平均绝对误差、对称平均绝对百分比误差和拟合优度3种评价指标下也均优于其他现有的拟合模型,具有良好的预测性能。展开更多
文摘Quantitative headspace analysis of volatiles emitted by plants or any other living organisms in chemical ecology studies generates large multidimensional data that require extensive mining and refining to extract useful information. More often the number of variables and the quantified volatile compounds exceed the number of observations or samples and hence many traditional statistical analysis methods become inefficient. Here, we employed machine learning algorithm, random forest (RF) in combination with distance-based procedure, similarity percentage (SIMPER) as preprocessing steps to reduce the data dimensionality in the chemical profiles of volatiles from three African nightshade plant species before subjecting the data to non-metric multidimensional scaling (NMDS). In addition, non-parametric methods namely permutational multivariate analysis of variance (PERMANOVA) and analysis of similarities (ANOSIM) were applied to test hypothesis of differences among the African nightshade species based on the volatiles profiles and ascertain the patterns revealed by NMDS plots. Our results revealed that there were significant differences among the African nightshade species when the data’s dimension was reduced using RF variable importance and SIMPER, as also supported by NMDS plots that showed S. scabrum being separated from S. villosum and S. sarrachoides based on the reduced data variables. The novelty of our work is on the merits of using data reduction techniques to successfully reveal differences in groups which could have otherwise not been the case if the analysis were performed on the entire original data matrix characterized by small samples. The R code used in the analysis has been shared herein for interested researchers to customise it for their own data of similar nature.
基金the grants from National Tackle Key Science and Technology Program sduring the Ninth Five-Year Plan Period, No.96-903-01-11the grants from State Administration of Traditional Chinese Medicine of People's Republic of China,No.00-01LP16
文摘BACKGROUND: Convalescence is an important stage of stroke treatment. A lot of patients have somatic and mental disorders at various degrees. The primary standard can only reflect partial conditions of somatic disorder; in addition, multiple dimensions of patients at the phase of stroke convalescence are further observed by using a lot of standards, such as signs and symptoms of traditional Chinese medicine, daily activity and psychological status. OBJECTIVE: To analyze the outcome assessments of the cases of stroke convalescence measured with different criteria consisting of various dimensions by a cross-sectional investigation of the condition of stroke convalescent patients. DESIGN: Scale evaluation. SETTING: Departments of Clinical Epidemiology Exploratory Development and Neurology, the Second Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine; National Center for Training of Design, Measurement and Evaluation in Clinical Research,Guangzhou University of Traditional Chinese Medicine. PARTICIPANTS: A total of 194 stroke convalescent patients treated in the Second Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine from July 26, 2000 to February 28, 2001 were taken as subjects of the study. There were 126 males and 68 females aged from 40 to 89 years, and the illness course ranged from 14 to 181 days. All patients met diagnosis-treatment criteria of stroke (the second version)[DTCS(V2.0)] and various kinds of diagnostic points of cerebrovascular diseases; moreover, all patients provided confirmed consents. METHODS: They were assessed by assessment methods including the following assessment instruments: DTCS(V2.0), self-designed scale of traditional Chinese medicine (TCM) symptoms (28 symptoms and physical signs were scored as 0, 1, 2 marks from none to severity), modified Edinburgh-Scandinavia stroke scale (a total of 45 marks, 0 to 15 marks as mild defect, 16 to 30 as moderate defect, 31 to 45 as severe defect), modified Barthel activities of daily life (ADL) index (a total of 100 marks, less than 60 marks as unable self-care), vitality and mental health (subscales derived from Health Survey Questionnaire, SF-36). The collected data from scales and inter-scale correlation were processed by the statistic methods mainly including descriptive analysis, Spearmen correlation analysis, factor analysis, etc. MAIN OUTCOME MEASURES: ① Average scores of scales and criteria; ② correlation between modified Edinburgh-Scandinavia stroke scale and other scales. RESULTS: All of the patients completed the assessment, and analyzed in the final analysis. ① The average scores of the scales and criteria: The average scores of DTCS(V2.0), self-designed scale of TCM symptoms, modified Edinburgh-Scandinavia stroke scale, modified Barthel ADL index, vitality and mental health scales were 6.51±6.29, 13.73±6.97, 7.56±7.35, 63.58±23.68, 52.79±23.32 and 62.83±22.75 respectively. ② Correlation between modified Edinburgh-Scandinavia stroke scale and other scales: The Spearman correlation coefficients (R ’) of modified Edinburgh-Scandinavia stroke scale with diagnosis-treatment criteria of stroke, scales of TCM symptoms, modified Barthel ADL index, vitality scale and mental health scale were 20.885, 0.302, -0.824, -0.294 and -0.258 respectively. CONCLUSION: The modified Edinburgh-Scandinavia stroke scale and DTCS(V2.0) shared the same assessment dimension, so they can be mutually alternated in some clinical practices. Discrepancy in measurements of health status was gained due to the diverse dimensions applied in outcome assessments. It is necessary to build up a multi-dimensional assessment criteria system, such as signs and symptoms, daily activities and psychological status, for assessing the stroke convalescent cases in a more comprehensive scope and reflecting the efficacy of TCM treatment scientifically.
文摘针对液晶显示器(LCD)面板的“Chip/FPC on Glass”(C/FOG)工艺生产制造过程中存在的计量延迟大、生产异常无法提前预测的问题,本文提出一种基于神经网络的C/FOG工艺生产制造虚拟计量方法。该方法利用生产机台上的传感器采集生产过程中的过程状态数据,构建基于多尺度一维卷积及通道注意力模型(MS1DC-CA)的虚拟计量模型。通过多个尺度的卷积核提取不同尺度范围内的状态数据特征。在对含有缺失值的原始数据预处理中,提出了基于粒子群算法改进的K近邻填补方法(PSO-KNN Imputation)进行缺失值填充,保留特征的同时,减少因填充值引入的干扰。最后在实际生产采集的数据上进行实验对比分析,实际不良率主要集中在0.1%~0.5%,该虚拟计量模型的拟合均方误差为0.397 7‱,低于其他现有拟合模型,在平均绝对误差、对称平均绝对百分比误差和拟合优度3种评价指标下也均优于其他现有的拟合模型,具有良好的预测性能。
