Prediction of Phishing Sites in Network using Naive Bayes compared over Random Forest with improved Accuracy

Citations Over TimeTop 10% of 2023 papers

Abstract

The main aim of this research article is to improve the accuracy rate in efficient Novel prediction of phishing sites in the network by using Naive Bayes (NB) in comparison with Random Forest (RF) Classifier. Materials & Methods: This research makes use of the publicly accessible Kaggle data set for efficient novel network prediction of phishing sites. The sample size for efficient Innovative prediction of phishing sites in the network with an enhanced accuracy rate was 80 (Group 1 = 40 and Group 2 = 40), and calculations were done using G-power 0.8 with alpha and beta values of 0.05 and 0.2, respectively, and a 95% confidence interval. Naive Bayes (NB) with N=10 samples and Random Forest (RF) with N=10 samples efficiently forecast phishing sites in a network with an enhanced accuracy rate. Results: The Naive Bayes (NB) classifier has 95.58 percent greater accuracy rates than Random Forest (RF), which is 94.675 percent. The statistical significance level of this study is p0.05, or p=0.0481. Conclusion: Naive Bayes (NB) is more accurate than Random Forest (RF) at predicting novel phishing sites in the network.

Related Papers

• → Neural network Bayes error estimation(1994)5 cited
• → A Bayes Empirical Bayes Decision Rule for Classification(2005)5 cited
• → Bayes empirical Bayes approach to unsupervised learning of parameters in pattern recognition(2000)2 cited
• → Bayes Empirical Bayes Classification of Components Using Masked Data(2010)1 cited
• → <title>Nonparametric Bayes error estimation for HRR target identification</title>(1994)2 cited