Figure 1: Effect of total cholesterol and LDL on Fasting Blood Sugar in surveyed Indian population. | Scientific Reports

Figure 1: Effect of total cholesterol and LDL on Fasting Blood Sugar in surveyed Indian population.

From: Restoring Mitochondrial Function: A Small Molecule-mediated Approach to Enhance Glucose Stimulated Insulin Secretion in Cholesterol Accumulated Pancreatic beta cells

Figure 1

(A) Architecture of parameter optimization for the support vector regression machines. The number represents the steps in the flow diagram. Dotted arrow represents the normalized data which is used by cross validation process along with instantiated SVR machine. Straight arrow designates the sequence flow of the experiment. (B) Plot of RMSE for a linear nu-SVR with other parameters remaining constant, X-Axis: [values of nu], Y-Axis: Cost (C), Z-Axis: RMSE (FS). (C) Plot of RMSE (FS) described as a function of nu SVR with other parameters [Kernel Type, Epsilon, Cost]. X-Axis: Nu, Y-Axis: RMSE (FS), Color Code: Epsilon, Size (0–100): Cost (C), Shape: Kernel function; circular: Linear Kernel; square: Radial Basis Kernel (RBF). (D) Plot of RMSE (FS) generated through LOO cross validation using Evolutionary Computational Approach search technique. X axis: SVM-Cost (C); Y-Axis: RMSE (FS), Color Code: Epsilon, Size: Nu. (E) Optimal coefficients of linear nu SVR describing the slope of the optimal hyperplane against each variable (weights) ; normalized relative weights of the predictors including the rank as expressed in the charts (E(i)) as well as histogram ((E(ii)) depicting relative impact of predictors on fasting blood sugar (FS).

Back to article page