Score Function
Scoring functionis the type of "score" available for the selected model, such as the predicted value of the target, the probability of the predicted value, the probability of the selected target value.
Scoring function type
- Predicted value: the predicted value of the target outcome;
- Probability of predicted value: correct value expressed as a proportion;
- Probability of the selected value:from 「 Value columnmiddleSelect a value from the drop-down list,Available values are defined by the model.
- Confidence: A probability measure associated with the predicted value of a categorical target. For Binary Logistic Regression, Multinomial Logistic Regression, and Naive Bayes models, the result is the same as the probability of the predicted value. For Tree and Ruleset models, the result is always less than the probability of the predicted value.
- Node number: the predicted terminal node number of the tree model;
- Standard error: standard error of the predicted value;
- Cumulative Hazard: Estimates the cumulative hazard function, which indicates the probability of observing an event at or before a specified time, given the values of the predictor variables.
- Nearest neighbor element: the ID of the nearest neighbor element, which belongs to the value of the case label variable;
- Kth nearest neighbor: The ID of the Kth nearest neighbor. Enter an integer in the Value column as the k value, which is the value of the case label variable.
- Distance to nearest neighbor: Depending on the model, Euclidean or city block distance is used;
- Distance to the Kth nearest neighbor: Enter an integer in the Value column as the value of k. Depending on the model, either the Euclidean or city block distance will be used.
Classification of scoring functions
- Bayesian Scoring Function
- Scoring function based on information theory