InfoDecipher | Innovative Predictive Modeling

References and Links

Definition of Predictive Modeling - Predictive modeling is used to create statistical models of future behavior. This is used in the area of data mining concerned with forecasting probabilities and trends. A predictive model is made up of a number of predictors--variable factors that are likely to influence future behavior or results. In marketing, for example, a customer's gender, age, and purchase history might predict the likelihood of a future sale.

Modeling Process - To create a predictive model, data is collected for the relevant predictors, a statistical model is formulated, predictions are made and the model is validated (or revised) as additional data becomes available. The model may employ various algorithms.

Logistic Regression - This is one of the commonly used statistical algorithms in predictive modeling. It is particularly effective for binary target variables (such as response/non-response, default/non-default). Check here for more technical details. Within logistic regression, there are variations to produce prediction. Commonly used methods may be classified into the 3 general categories

- Linear Logistic Regression

- Parametric Non-linear Logistic Regression

- Multi-layer Perception (MLP)