Data Quality Issues for Speech Emotion Recognition System Using Novel Logistic Regression Algorithm and Compare Accuracy with Random Forest Algorithm

To compare the accuracy of the Random Forest algorithm with the Novel Logistic Regression Technique in order to predict the data quality issues for the voice emotion identification system. The Speech Emotion Recognition System compares its machine learning techniques with the accuracy of its metrics. To create a procedure to predict speech emotion and to obtain the best accuracy to compare the algorithm, gradient boosting builds upon the tree and random forest algorithms. The algorithm must to be sufficiently effective to generate the precise speech emotion prior to the speech commencing. There are 20 samples total for the two groups, and the sample size taken into account for carrying out the task using a pretest power analysis was 80%. For data quality concerns, the Random Forest Algorithm and the Novelistic Regression algorithm were used, with accuracy rates of 90.% and 87.9%, respectively. A statistically significant difference in accuracy between two algorithms is found to be 1.2 (p0.05) using independent samples t-tests. Compared to the random forest technique, the new logistic regression approach performs noticeably better.