What is predictive analytics (PA)?

• Predictive analytics (PA) (or predictive modeling) is the use of historical data to forecast future outcomes.

• The complexity of PA may vary. For example, PA may rely on relatively simple approaches, such as models with a small number of measures (i.e. predictors, variables). Alternatively, in settings with large amounts of data, they may involve complex machine-learning algorithms.

• PA may make predictions of binary outcomes, which take just two values such as yes or no. For example, we can predict whether students will graduate from high school on time or not. With binary outcomes, predictive analytics can also be used to estimate the probability of the outcome occurring. For example, we can predict the probability that a student will graduate from high school. This probability is between 0 and 1.

• PA may also forecast continuous outcomes. For example, we can predict the grade point average (GPA) students will achieve.

• To narrow the scope of the class, we will focus only on binary outcomes.