Automatic Identification of Fake News Circulation in Social Media using Logistic Regression over Naïve Bayes and Xg Boost Algorithm to Improve Accuracy
Keywords:Fake news detection, Machine learning, Novel logistic regression, Xgboost, Logistic regression, Supervised algorithm.
Aim: To Detect the Fake News in Social Media using Logistic Regression and XGBoost Algorithms. To achieve accuracy a novel logistic regression is used. Materials and Methods: The datasets extracted from the kaggle data world and those datasets named as ‘TRUE’ and ‘FAKE’. Accuracy and loss are performed with datasets from kaggle library. The total sample size is 20. The two groups considered were Logistic regression (N=10) and xg boost (N=10). Results: Novel logistic regression pops up with the mean accuracy of when contrasted with the Xgboost algorithm. Mean Accuracy value for Logistic Regression and XgBoost Algorithm is 93.68 and 92.93 respectively. Mean Loss value for Logistic Regression and Xg boost is 6.31 and 7.06 respectively. Ultimately Novel logistic regression(NLR) pops up with a better significant value than the Xgboost algorithm. The two algorithms NLR and Xgboost are statistically satisfied with the independent sample T-Test value (p<0.006) with confidence level of 95%. Conclusion: Detecting fake news significantly seems to be better in Logistic Regression than xg boost.