Predictive analytics' techniques
There are different types of predictive analytics techniques based on the method of analysis. The nature of the learning input and predicted outcome remains the same for all the techniques, which are explained in the sections that follow.
Regression-based prediction
Regression-based prediction is predicting the unknown value of a variable from the known value of another variable. We predict scores on one variable based on the scores of a second (or more) variable.
There are two methods of regression that are based on the outcome/target variable:
Linear regression
Logistic regression
An example of regression-based prediction is Stochastic Gradient Descent (SGD).
Model-based prediction
In model-based prediction, we assume that the data follows a specific statistic model. By following this method, we can take advantage of the structure of the data by building a parametric model for a given data distribution.
An example of model-based prediction is Naïve Bayes.
