Predictive Modeling
Your data and the kind of analysis you need to do determines the type of prediction to apply. Spotfire provides four types of predictive modeling tools.
| Predictive Model | Description |
|---|---|
| Linear regression | Models the numeric response column as a weighted sum of the predictor columns. The weights, also known as the regression coefficients, are selected by the method of least squares, which minimizes the sum of the squared differences between the observed response and the predictions based on the weighted sum. |
| Regression tree | A nonparametric regression method that creates a binary tree by recursively splitting the data on the predictor values. The splits are selected so that the variation in the Response column is smaller in each child node than in the parent node. Various options are used to control how deep the tree is grown. Predictions are based on the mean of all the Response values in the terminal node for an observation. |
| Logistic regression | A classification method used when the Response column is categorical with only two possible values. The probability of the possible outcomes is modeled with a logistic transformation as a weighted sum of the Predictor columns. The weights or regression coefficients are selected to maximize the likelihood of the observed data. |
| Classification tree | A nonparametric classification method that creates a binary tree by recursively splitting the data on the predictor values. The splits are selected so that the variation in the Response column is smaller in each child mode than in the parent node. Various options are used to control how deep the tree is grown. Predictions are based on the mean of all the Response values in the terminal node for an observation. |