The regression feature of Predict allows you to view your data and address the relationships between value points. Predict can perform 3 regression types.
Predict can perform:
- Logistic Regression Analysis
- Logistic regression analysis models variables with a set number of parameters depending on the quantity of input values. This type of regression produces categorial predictions, depending on your data. This method depends on the outcome you’re trying to predict. You can predict any continuous (how many/how much/how long) or binary (yes/no, 1/0) outcome, and Predict will recognize the best modeling method for either possibility.
- Ordinary Least Square Analysis (OLS)
- The OLS analysis method minimizes the sum of square differences between your observed and predicted values. This allows for a clearer variable comparison in your algorithm.
- Decision Tree – Chi-square Automatic Interaction Detection (CHAID)
- This type of model builds a tree to identify relationships between variables. With CHAID, you will learn how variables best merge to explain outcomes. This model provides a visual representation of the relationships in your data set.