DECISION MAKING BY VARIABLE CONTRIBUTION IN DISCRIMINANT, LOGIT, AND REGRESSION ANALYSES

2004 ◽  
Vol 03 (02) ◽  
pp. 265-279 ◽  
Author(s):  
STAN LIPOVETSKY ◽  
MICHAEL CONKLIN
Keyword(s):  
Decision Making ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Statistical Models ◽  
Regression Analyses ◽  
Discriminant Functions ◽  
Multivariate Statistical ◽  
Linear Discriminant ◽  
Variable Contribution ◽  
Linear Aggregate

Comparative contribution of predictors in multivariate statistical models is widely used for decision making on the importance of the variables for the aims of analysis and prediction. However, the analysis can be made difficult because of the predictors' multicollinearity that distorts estimates for coefficients in the linear aggregate. To solve the problem of the robust evaluation of the predictors' contribution, we apply the Shapley Value regression analysis that provides consistent results in the presence of multicollinearity both for regression and discriminant functions. We also show how the linear discriminant function can be constructed as a multiple regression, and how the logistic regression can be approximated by linear regression that helps to obtain the variables contribution in the linear aggregate.

The alternative negative constructions in spoken and written Korean: Logistic regression analysis

2019 ◽  
Vol 15 (2) ◽  
pp. 419-442
Author(s):  
Beom-mo Kang
Keyword(s):  
Logistic Regression ◽  
Regression Analysis ◽  
Logistic Regression Analysis ◽  
Mixed Effects ◽  
Written Language ◽  
Regression Analyses ◽  
Relevant Factor ◽  
Factors Influencing ◽  
Frequency Of Use ◽  
Relevant Factors

AbstractAdopting quantitative corpus-based methods, this paper focuses on the alternative negative constructions in Korean, [anV] and [Vanhda]. Logistic regression analyses for a mixed-effects model were carried out on data drawn from the Sejong Korean Corpus. Certain features of the verb or adjective in negative constructions significantly affect the use of the two negative constructions. A relevant factor is register/medium (spoken or written), among other significant interactions of factors. Furthermore, the fact that frequency is consistent with other relevant factors, together with certain diachronic facts of Korean, supports the claim that frequency of use plays an important role in linguistic changes. Another finding is that, notwithstanding noticeable differences between spoken and written language, the factors influencing the use of the two negative constructions in Korean are largely similar in the spoken and written registers.

Combined serum amylase and lipase determinations for diagnosis of suspected acute pancreatitis

1993 ◽  
Vol 39 (12) ◽  
pp. 2495-2499 ◽  
Author(s):  
J P Corsetti ◽  
C Cox ◽  
T J Schulz ◽  
D A Arvan
Keyword(s):  
Acute Pancreatitis ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Test Performance ◽  
Serum Amylase ◽  
Discriminant Functions ◽  
Clinical Implementation ◽  
Sensitivity Specificity ◽  
Bivariate Approach

Abstract Serum amylase and lipase measurements are often used to diagnose acute pancreatitis. This study addresses the question of whether it is advantageous to order serum amylase and lipase tests simultaneously. We evaluated performance of the two tests separately and in combination through a retrospective study of patients for whom both amylase and lipase determinations were ordered. Initial analysis of test performance was conducted with a uniformly applied criterion based on determination of optimal sensitivity-specificity pairs. Individual tests and combinations of tests, including the "AND" and "OR" rules and discriminant functions, were examined. Only the discriminant approach demonstrated better performance than the lipase test alone. This finding was subsequently confirmed by logistic regression analysis. We conclude that ordering both tests simultaneously can be advantageous in diagnosing acute pancreatitis when a bivariate approach is used; however, this must be weighed against the difficulties associated with clinical implementation of such approaches.

scholarly journals Opening the Black Box: the Relationship between Neural Networks and Linear Discriminant Functions

1997 ◽  
Vol 14 (1) ◽  
pp. 19-30 ◽  
Author(s):  
Roger A. Kemp ◽  
Calum MacAulay ◽  
Branko Palcic
Keyword(s):  
Neural Networks ◽  
Decision Making ◽  
Classification Problem ◽  
Black Box ◽  
Discriminant Functions ◽  
Classification Problems ◽  
Statistical Decision ◽  
Feed Forward ◽  
Linear Discriminant ◽  
Feed Forward Neural Networks

Over the last ten years feed‐forward neural networks have become a popular tool for statistical decision making. During this time, they have been applied in many fields, including cytological classification. Neural networks are often treated as a black box, whose inner workings are concealed from the researcher. This is unfortunate, since the inner workings of a neural network can be understood in a manner similar to that of a linear discriminant function, which is the standard tool that researchers use for decision making.This paper discusses feed‐forward neural networks and some methods to improve their performance for classification problems. Their relationship to discriminant functions will be examined for a simple two‐dimensional classification problem.

scholarly journals Reasons and Factors Contributing to Chinese Patients’ Preference for Ustekinumab in Crohn’s Disease: A Multicenter Cross-Sectional Study

2021 ◽  
Vol 12 ◽  
Author(s):  
Lingya Yao ◽  
Xiao Zhu ◽  
Bule Shao ◽  
Rongbei Liu ◽  
Zhilun Li ◽  
...  
Keyword(s):  
Decision Making ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Crohn’S Disease ◽  
Shared Decision Making ◽  
Logistic Regression Analysis ◽  
Chinese Patients ◽  
Shared Decision ◽  
Contributing Factors ◽  
Cross Sectional

Background and Aims: Ustekinumab (UST) was approved in China for treating moderate-to-severe Crohn’s disease (CD) in 2020. We aimed to identify the reasons and possible contributing factors for UST preference in Chinese patients with CD.Methods: We conducted a multicenter cross-sectional survey among patients with moderate to severe CD who underwent UST treatment in 27 hospitals. Patients completed a 46-item questionnaire that included information on demographics, clinical characteristics, reasons in favor of UST and shared decision-making perception. Logistic regression analysis was performed to examine the predictive factors of different UST preferences.Results: Overall, 127 patients (73 males; mean age, 25.9 ± 9.9 years) completed the questionnaire. Most patients (74.8%) had biologic failure. The most common reason for the latest treatment disconnection was unresponsiveness to the previous medications. The major UST information sources were physicians (96.1%). Nearly half of the patients (44.9%) reported shared decision making regarding UST treatment. No difference was found in the decision-making patterns in terms of sex and age. The most influential reason for UST preference was “effectiveness” (77%, 98/127), followed by “safety” (65%, 83/127), “frequency of administration” (39%, 49/127), and “mode of administration” (37%, 47/127). Multivariate logistic regression analysis revealed that a positive self-rated health status was a contributing factor for UST preference with a low frequency of administration.Conclusion: This is the first multicenter survey of Chinese patients with CD to identify the possible contributing factors for UST preference. Treatment choice should be discussed with patients because individual preferences are determined by diverse factors.

scholarly journals Analysis of Nonlinear Relationships in Dual Epidemics, and Its Application to the Management of Grapevine Downy and Powdery Mildews

2009 ◽  
Vol 99 (8) ◽  
pp. 930-942 ◽  
Author(s):  
Serge Savary ◽  
Lionel Delbac ◽  
Amélie Rochas ◽  
Guillaume Taisant ◽  
Laetitia Willocquet
Keyword(s):  
Decision Making ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Powdery Mildew ◽  
Disease Control ◽  
Probabilistic Models ◽  
Nonlinear Modeling ◽  
Powdery Mildews ◽  
Cropping Season ◽  
Tactical Decision

Dual epidemics are defined as epidemics developing on two or several plant organs in the course of a cropping season. Agricultural pathosystems where such epidemics develop are often very important, because the harvestable part is one of the organs affected. These epidemics also are often difficult to manage, because the linkage between epidemiological components occurring on different organs is poorly understood, and because prediction of the risk toward the harvestable organs is difficult. In the case of downy mildew (DM) and powdery mildew (PM) of grapevine, nonlinear modeling and logistic regression indicated nonlinearity in the foliage–cluster relationships. Nonlinear modeling enabled the parameterization of a transmission coefficient that numerically links the two components, leaves and clusters, in DM and PM epidemics. Logistic regression analysis yielded a series of probabilistic models that enabled predicting preset levels of cluster infection risks based on DM and PM severities on the foliage at successive crop stages. The usefulness of this framework for tactical decision-making for disease control is discussed.

Retirement Satisfaction for Retirees and their Spouses

2004 ◽  
Vol 25 (2) ◽  
pp. 262-285 ◽  
Author(s):  
Deborah B. Smith ◽  
Phyllis Moen
Keyword(s):  
Decision Making ◽  
Logistic Regression ◽  
Life Course ◽  
Well Being ◽  
Life Course Perspective ◽  
Regression Analyses ◽  
Multivariate Logistic Regression ◽  
Retirement Decision ◽  
Joint Retirement ◽  
Spousal Influence

This study investigates factors related to retirees’ and their spouses’ individual and joint retirement satisfaction using decision-making theory and a life course perspective. The sample includes 421 retired respondents (ages 50 to 72) and spouses from the Cornell Retirement and Well-Being Study. Although 77% of retirees report retirement satisfaction, only 67% of their spouses are satisfied; even fewer couples (59%) report joint satisfaction. Multivariate logistic regression analyses reveal that retirees’ and spouses’ individual and joint reports of retirement satisfaction are related to perceptions of spousal influence on the retirement decision, with effects varying by gender. Those couples most likely to report being satisfied with retirement, individually and jointly, are retired wives and their husbands where wives reported that their husbands were not influential in their retirement decision. The results underscore the importance of regarding retirement as gendered and as both an individual- and a couple-level experience.

scholarly journals Routine Health Information Use for Decision-making and Associated Factors among Managers Working at Public Hospitals in North Shewa Zone of Oromia Regional State, Ethiopia: A Mixed-Method Study

2021 ◽  
Author(s):  
Gemechu Tulu ◽  
Takele Gezahegn Demie ◽  
Tesfalem Teshome
Keyword(s):  
Decision Making ◽  
Information System ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Health Information ◽  
Associated Factors ◽  
Health Information System ◽  
Information Use ◽  
Public Hospitals ◽  
Supportive Supervision

Abstract Background: Routine health information systems (RHIS) are vital for the acquisition of data for health sector planning, monitoring, and evaluation, patient management, health education, resource allocation, disease prioritization, and decision-making. Routine health information use for decision-making is low in Ethiopia. Thus, the study aimed to assess level of routine health information use and its associated factors among managers working at public hospitals in North Shewa, Ethiopia.Methods: A facility-based cross-sectional study design with both quantitative and qualitative data collection methods was conducted from May to June 2020. A total of 102 randomly selected managers working in public hospitals in North Shewa were included in the quantitative study while six in-depth interviews were performed for the qualitative method. Quantitative data were collected using a structured self-administered questionnaire and interview guide by trained data collectors, cleaned, coded, and entered into Epi-info version 7.1 software and transferred into SPSS version 23 software for Window for further statistical analysis. Both bivariable and multivariable logistic regression analyses were performed. In the multiple logistic regression analysis, a less than 0.05 P-value was considered statistically significant. The odds ratio along with a 95% confidence interval (CI) were estimated to measure the strength of the association. Thematic analysis was done for key informant interview data.Result: In this study, the level of routine health information use for decision-making was 71.6% (95% CI: 61.8%, 79.4%). According to the multivariable logistic regression analysis, training on health information system (AOR = 0.28, 95% CI: 0.08-0.98) and supportive supervision (AOR = 0.27, 95% CI: 0.09-0.78) were found significantly associated with the use of routine health information for decision-making. Moreover, the lack of staff motivation and computer and data analysis skills were the major reasons for not using routine health information.Conclusions: Three-fourth of the managers working at public hospitals used routine health information for decision-making. Training on health information system and supportive supervision were factors associated with the use of routine health information. Therefore, training of managers and the provision of supportive supervision were highly recommended to improve the use of routine health information managers for decision-making at public health institutions.

scholarly journals Development and Internal Validation of Supervised Machine Learning Algorithms for Predicting the Risk of Surgical Site Infection Following Minimally Invasive Transforaminal Lumbar Interbody Fusion

2021 ◽  
Vol 8 ◽  
Author(s):  
Haosheng Wang ◽  
Tingting Fan ◽  
Bo Yang ◽  
Qiang Lin ◽  
Wenle Li ◽  
...  
Keyword(s):  
Machine Learning ◽  
Decision Making ◽  
Logistic Regression ◽  
Interbody Fusion ◽  
Clinical Decision Making ◽  
Transforaminal Lumbar Interbody Fusion ◽  
Clinical Decision ◽  
Stepwise Logistic Regression ◽  
Lumbar Interbody Fusion ◽  
Regression Analyses

Purpose: Machine Learning (ML) is rapidly growing in capability and is increasingly applied to model outcomes and complications in medicine. Surgical site infections (SSI) are a common post-operative complication in spinal surgery. This study aimed to develop and validate supervised ML algorithms for predicting the risk of SSI following minimally invasive transforaminal lumbar interbody fusion (MIS-TLIF).Methods: This single-central retrospective study included a total of 705 cases between May 2012 and October 2019. Data of patients who underwent MIS-TLIF was extracted by the electronic medical record system. The patient's clinical characteristics, surgery-related parameters, and routine laboratory tests were collected. Stepwise logistic regression analyses were used to screen and identify potential predictors for SSI. Then, these factors were imported into six ML algorithms, including k-Nearest Neighbor (KNN), Decision Tree (DT), Support Vector Machine (SVM), Random Forest (RF), Multi-Layer Perceptron (MLP), and Naïve Bayes (NB), to develop a prediction model for predicting the risk of SSI following MIS-TLIF under Quadrant channel. During the training process, 10-fold cross-validation was used for validation. Indices like the area under the receiver operating characteristic (AUC), sensitivity, specificity, and accuracy (ACC) were reported to test the performance of ML models.Results: Among the 705 patients, SSI occurred in 33 patients (4.68%). The stepwise logistic regression analyses showed that pre-operative glycated hemoglobin A1c (HbA1c), estimated blood loss (EBL), pre-operative albumin, body mass index (BMI), and age were potential predictors of SSI. In predicting SSI, six ML models posted an average AUC of 0.60–0.80 and an ACC of 0.80–0.95, with the NB model standing out, registering an average AUC and an ACC of 0.78 and 0.90. Then, the feature importance of the NB model was reported.Conclusions: ML algorithms are impressive tools in clinical decision-making, which can achieve satisfactory prediction of SSI with the NB model performing the best. The NB model may help access the risk of SSI following MIS-TLIF and facilitate clinical decision-making. However, future external validation is needed.

Sign in / Sign up

Export Citation Format

Share Document