Browsing by Author "Narayanapillai Sreekumaran Nair"

Now showing 1 - 2 of 2

Classification Accuracy of Linear Discriminant Function using Principal Components with Multiple Correlated Variables: A Simulation based Exploration
(Wolters Kluwer Medknow Publications, 2025) Akash Mishra; Sachit Ganapathy; Narayanapillai Sreekumaran Nair; Durgesh Shukla; Ashish Kumar Yadav; Rajaat Vohra; Kuldeep Soni
Background: Linear Discriminant Analysis (LDA) is a powerful and widely used technique for classification with correlated variables. Principal Components (PCs) group these variables into linear combinations and produce independent variables. The LDA on these PC’s may provide better classification accuracy in clinical diagnostics than on usual measurements. Methodology: Two datasets were utilized for demonstration: one from a Sudden Sensorineural Hearing Loss (SSNHL) case-control study and the other from a Gall Bladder (GB) case-control study. Linear Discriminant Analysis (LDA) was conducted on the actual correlated measured variables for group classification, as well as on the derived principal component variables, to compare their classification accuracies. Performance metrics including Sensitivity, Specificity, Positive Predictive Value (PPV), Negative Predictive Value (NPV), Classification Accuracy, and F1 Score were assessed. For validation, a third simulated dataset was employed. Additionally, LDA was performed on each dataset using eigenvectors of the control group applied to the cases and vice versa, revealing a strong agreement in classification as measured by the kappa statistic. Results: When LDA was applied to the actual lipid measurements in the SSNHL dataset, the classification accuracy was 57.2%, and the F1 score was 39.7%. However, when LDA was performed using principal components (PCs), the classification accuracy markedly improved to 99.2%, with an F1 score of 98.5%. Similarly, for the GB cancer dataset, the classification accuracy and F1 score were initially 77.2% and 77.3%, respectively. Upon applying LDA with the PCs, these metrics were significantly enhanced to 98.4% and 98.3%, respectively. For the simulated dataset, both the classification accuracy and F1 score were 99.1%. The study also demonstrated that the classification accuracy and F1 score remained consistent regardless of whether the eigenvectors from the cases or controls were used to classify new subjects (Kappa Statistic = 0.962, P < 0.001). Conclusion: In group separation, utilizing principal components significantly improves classification accuracy and overall performance metrics, outperforming the use of the original correlated predictors. © 2025 Medical Journal of Dr. D.Y. Patil Vidyapeeth.
Identifying the clinical relative importance of each correlated outcome variables in multivariate approach: an exploration using ACCORD trial data
(Taylor and Francis Ltd., 2025) Akash Mishra; Narayanapillai Sreekumaran Nair; Kottenyen Thazhath Harichandrakumar; Binu Vs; Santhosh Satheesh
In scenarios involving correlated endpoints, multivariate methods offer increased robustness for comparisons. However, understanding the individual contribution of each variable toward multivariate hypothesis rejection remains underexplored. Usually, this question is sidelined, and separate univariate analyses are performed. This paper addresses this gap by demonstrating the relative importance and contribution of variables toward the rejection of multivariate hypotheses, comparing it against a univariate approach using clinical trial data. Using the ACCORD lipid trial dataset, which includes lipid measurements of triglycerides (TG), low-density lipoprotein (LDL), and high-density lipoprotein (HDL), we employed Hotelling’s T2 multivariate statistic for two-group comparisons. We showcased the significance and relative importance of contributions through standardized discriminant function coefficients and partial F tests. Additionally, we investigated the impact of varying correlation levels on the significance of each variable’s contribution in multivariate versus univariate approaches. Our results revealed significant lipid differences in a multivariate context at the 12th and 36th months. Across both follow-ups, TG exhibited the highest relative importance and contribution, followed by HDL and LDL. Notably, in the 36th month, the univariate approach rendered LDL’s contribution insignificant for group separation, contrasting with the significant contribution identified in the multivariate approach. Furthermore, the significance likelihood of variable contributions in group separation within the multivariate approach increased with rising correlation levels. The simulation technique and the power analysis was also adopted to characterize the features of the proposed method. Our approach enables the evaluation of the relative importance and significance of each variable’s contribution within the multivariate framework. This methodology holds promise for enhancing the interpretation of clinical trial analysis outcomes, particularly when dealing with multiple correlated endpoints. © 2025 Taylor & Francis Group, LLC.