A comparison of logistic regression and the AlexNet algorithm for the accuracy analysis of fallopian tube tumor prediction.

The main aim of this study is to compare & determine the accuracy to predict tumor on a fallopian tube based on colposcopy images among AlexNet and logistic regression. Using ClincalC and the parameters of G-power for each group of 0.8, alpha value of 0.05, and beta value of 0.2, the total number of samples (colposcopy images) is calculated as 5106 (sample size of the group 1 of 2553 and group 2 of 2553). In order to improve results, this dataset from Kaggle is subjected to 15 iterations for each group. In order to process the image and determine the disease stage, the AlexNet and logistic regression techniques are used in MATLAB. The results of the SPSS statistical analysis tools are used to select the most suitable approach. In comparison to logistic regression, which had an accuracy of 90.01%, the results showed that AlexNet had a greater accuracy of 93.31%. The independent t-test value is 0.000, which is below the acceptable limit (0.05). This shows the groups are highly significant to each other. It is obtained that the suggested AlexNet method predictions were more precise than the logistic regression model. [ABSTRACT FROM AUTHOR]