Medical decision making : an international journal of the Society for Medical Decision Making
Jan 1, 2016
OBJECTIVE: To evaluate the impact of the synthetic minority oversampling technique (SMOTE) on the performance of probabilistic neural network (PNN), naïve Bayes (NB), and decision tree (DT) classifiers for predicting diabetes in a prospective cohort ...
Proceedings of the National Academy of Sciences of the United States of America
Apr 17, 2018
Word embeddings are a powerful machine-learning framework that represents each English word by a vector. The geometric relationship between these vectors captures meaningful semantic relationships between the corresponding words. In this paper, we de...