2.8 Logistic Regression Classification

In logistic regression, a supervised machine learning algorithm, a logistic function, also known as a sigmoid function, is used to produce a probability value between 0 and 1 based on inputs that are independent variables (see Fig 2H). For instance, Class 0 and Class 1 are the two classes. If the logistic function value of an input exceeds the threshold value of 0.5, it is classed as Class 1 or Class 0. It is called regression because it is essentially used for classification problems and is a continuation of linear regression [31].

Furthermore, the AUC-ROC curve was used to analyse the feature importance of each input parameter on the tensile strength. AUC-ROC analysis is a useful tool for assessing binary classification models’ performance and determining which ones are useful for a given task. Plot the ROC curve, with TPR on the y-axis and FPR on the x-axis. Each point on the curve represents a different threshold setting. Calculate the area under the ROC curve. A perfect classifier would have an AUC of 1. Compare the AUC values of different classifiers or different models to determine which one performs better. Different feature importance plots like ANOVA and Pearson’s heatmap were plotted. The acronym for analysis of variance is ANOVA. It is a statistical test that examines the statistical distinctions between the data’s numerical and categorical feature sets. Generally speaking, it looks for relationship patterns between the different data features. Analysing the variances of the samples taken from a population may also be used to test hypotheses by determining if two or more population means are equal. A statistical indicator of the linear relationship between two variables is the correlation coefficient, sometimes the Pearson correlation coefficient. The correlation coefficient is another name for the Pearson correlation. The application of the Pearson correlation coefficient spans numerous disciplines, including biology, sociology, and psychology. The Pearson correlation coefficient is frequently used in psychology to assess the extent to which two constructs—intelligence and academic achievement—are related. The Pearson correlation coefficient can be used in sociological studies to investigate the relationship between education level and income. Refer to Fig 2I, which depicts a person’s heatmap workflow.