Phishing Attack Detection using Logistic Regression and Novel Random Forest Algorithm to improve Accuracy

—The objective is to more accurately forecast phishing attacks that harvest sensitive data from unsuspecting users by utilizing Logistic Regression in comparison to the Novel Random Forest Algorithm. Two groups are used, such as the innovative Random Forest Algorithm and Logistic Regression. This suggested approach will be applied to a total of 110 photographs. Thirty percent (testing dataset) and one hundred fifty five images (training dataset) were utilized in this sample dataset.An experiment in programming was conducted using N=20 iterations to compare the new Random Forest Algorithm with Logistic Regression. The accuracy of the computations was checked once they were completed. Using an 80% G-Power value, SPSS was used to forecast the dataset’s significance value. The novel Random Forest algorithm has a much higher identification rate of 90.54 (p = 0.02) and a high accuracy for detecting phishing attacks. According to the descriptive statistical data, the Random Forest algorithm outperforms Logistic Regression in detecting phishing assaults.