Implementation of an Innovative Adaptive Logistic Regression Algorithm to Improve the Early Identification Rate of Seizure in Comparison with Naive Bayes Algorithm

V. Narendra Reddy ,R. Priyanka
Keywords: Seizure Detection, Naive Bayes, Innovative Adaptive Logistic Regression Algorithm, Machine Learning, Early Identification Rate, Classification Accuracy, Outbreak Prediction. ,

Abstract

Aim: The aim of the research is to improve the early identification rate of epileptic seizures using the Innovative Adaptive logistic regression algorithm in comparison with naive bayes algorithm. Methods and Materials: The total of 6426 samples are collected from the UCI repository. Group 1 represents the Innovative Adaptive logistic regression algorithm and group 2 represents the naive bayes algorithm. The G power calculation was done with 80% of power and alpha of 0.05. Each group with 20 samples were taken for SPSS analysis. Results: The Innovative Adaptive logistic regression algorithm has achieved the accuracy, Precision, Recall and Specificity of 97.0 %, 95 %, 90 %, and 89 % when compared to naive bayes algorithm with 78 %, 94 %, 78 % and 74 %. The logistic regression algorithm has achieved the significance of 0.043 (p < 0.05). Conclusion: In this study it is concluded that the logistic regression algorithm has significantly greater accuracy when compared with the naive bayes algorithm.