Performance Of Naive Bayes Research Articles

Naive Bayes Performance in Analysis of Public Opinion Sentiment Against COVID-19

The huge impact caused by the COVID-19 pandemic has made many people express their opinions on Twitter social media. There are various responses given by the community that are negative and positive. The dataset comes from kaggle with more than 750 tweets of data. Classification designed by the Naive Bayes method. Implementation through preprocessing, case folding, tokenizing, stopword removal, TF-IDF, and cross validation has been able to produce quite high accuracy. After classification, validation will be carried out with Cross Fold Validation. The best value is on cv5 where accuracy = 0.847, precision = 0.855, recall = 0.83, and f1 score = 0.842.

Bug Prioritization to Facilitate Bug Report Triage

The large number of new bug reports received in bug repositories of software systems makes their management a challenging task. Handling these reports manually is time consuming, and often results in delaying the resolution of important bugs. To address this issue, a recommender may be developed which automatically prioritizes the new bug reports. In this paper, we propose and evaluate a classification based approach to build such a recommender. We use the Naive Bayes and Support Vector Machine (SVM) classifiers, and present a comparison to evaluate which classifier performs better in terms of accuracy. Since a bug report contains both categorical and text features, another evaluation we perform is to determine the combination of features that better determines the priority of a bug. To evaluate the bug priority recommender, we use precision and recall measures and also propose two new measures, Nearest False Negatives (NFN) and Nearest False Positives (NFP), which provide insight into the results produced by precision and recall. Our findings are that the results of SVM are better than the Naive Bayes algorithm for text features, whereas for categorical features, Naive Bayes performance is better than SVM. The highest accuracy is achieved with SVM when categorical and text features are combined for training.

Improving Naïve Bayes Classifiers Using Interaction and Transformations Of Variables

An algorithm of generating functionally dependent variable transformations and interactions is used with random forest variable selection in conjunction with naive bayes is proposed. This algorithm improves performance of naive bayes substantively and is away for factoring in explicit variables to represent interactions between variables to overcome loss of information due to the conditional independence assumption. Adding interaction effects and transformations helped improve out of sample performance in 3 credit data sets by 5% but still did not match I* performance or logistic regression w tuned interactions. That makes sense as credit data is notorious for correlated variables which violate the independence assumption.