Logistic Probability Score
The logistic probability score function allows the user to obtain a predicted probability score of a given event using a logistic regression model. The logistic probability score works by specifying the dependent variable (binary target) and independent variables as input. The function runs the logistic model and fits the model and calculates the fitted probability of the event from the model.
If we are building a logistic regression model to predict a binary outcome (Y) using independent variables (X1, X2, X3, …, Xn) then the logistic model equation would look like,
Where p is the probability of the event (Y). From this equation, we can derive p-value as,
Application & Interpretation:
The function calculates the logistic probability score for the development sample. This function calculates the score for each observation present in the development dataset. If a given prospect has high probability score, we can interpret that the prospect has a higher inclination towards the overall objective function used in the model.
Input:
To run the Logistic Probability function in Analyttica TreasureHunt, you should select the target binary variable and one or many independent variables.
Output:
The function returns a probability score of each observation and appended at the end of the table.
See Also:
