Predictive Analytics and Data Mining

4.4. Naïve Bayesian

The data mining algorithms used for classification tasks are quite diverse. The objective of these algorithms is the same—prediction of a target variable. But, the method of predicting is drawn from a range of multidisciplinary techniques. The naïve Bayes algorithm finds its roots in statistics and probability theory. In general, classification techniques try to predict class labels based on attributes by best approximating the relationship between input and output variables. Every day, we mentally estimate a myriad of outcomes based on past evidence. Consider the process of guessing commuting time to work. First, commute time depends heavily on when you are traveling. If you are traveling during peak hours, most likely the ...