BayesiaLab 5.3: New Features & Updates (03/2014)
Here is a small selection of the new or updated features released in BayesiaLab 5.3:
- Redesigned: Contribution Analysis uses counterfactuals to measure the "added value" of individual driver on the target variable.
- Conditional Probability Trees can take advantage of contextual independencies in the parent-to-child relationships.
- Targeted Evaluation can now be performed with a limited number of questions.
- Markov Blanket Export in Visual Basic.
- Evidence Data Weighting for using the set of Evidence defined on the current network to compute a weight for each row of the data set.
- New missing values processing with Entropy-Based Imputations and Most Probable Explanation Imputations.
- Tree Discretization using Structural Coefficients.
- Search Function in Settings.