Cost-sensitive selective naive Bayes classifiers for predicting the increase of the h-index for scientific journals

Abstract

Machine learning community is not only interested in maximizing classification accuracy, but also in minimizing the distances between the actual and the predicted class. Some ideas, like the cost-sensitive learning approach, are proposed to face this problem. In this paper, we propose two greedy wrapper forward cost-sensitive selective naive Bayes approaches. Both approaches readjust the probability thresholds of each class to select the class with the minimum-expected cost. The first algorithm (CS-SNB-Accuracy) considers adding each variable to the model and measures the performance of the resulting model on the training data. The variable that most improves the accuracy, that is, the percentage of well classified instances between the readjusted class and actual class, is permanently added to the model. In contrast, the second algorithm (CS-SNB-Cost) considers adding variables that reduce the misclassification cost, that is, the distance between the readjusted class and actual class. We have tested our algorithms on the bibliometric indices prediction area. Considering the popularity of the well-known h-index, we have researched and built several prediction models to forecast the annual increase of the h-index for Neurosciences journals in a four-year time horizon. Results show that our approaches, particularly CS-SNB-Accuracy, achieved higher accuracy values than the analyzed cost-sensitive classifiers and Bayesian classifiers. Furthermore, we also noted that the CS-SNB-Cost always achieved a lower average cost than all analyzed cost-sensitive and cost-insensitive classifiers. These cost-sensitive selective naive Bayes approaches outperform the selective naive Bayes in terms of accuracy and average cost, so the cost-sensitive learning approach could be also applied in different probabilistic classification approaches.