Oracle Naive Bayes

Naive Bayes is a well-known algorithm for classification problems. The model is termed naïve because it treats all proposed prediction variables as being independent of one another. Naive Bayes is a fast, scalable algorithm that calculates conditional probabilities for combinations of attributes and the target attribute. From the training data, an independent probability is established. This probability gives the likelihood of each target class, given the occurrence of each value category from each input variable.

• Cross-validation is used to test model accuracy on the same data that were used to build the model. This is particularly useful when the number of cases available to build a model is small.
• The model output can be browsed in a matrix format. The numbers in the matrix are conditional probabilities that relate the predicted classes (columns) and predictor variable-value combinations (rows).