BayesiaLab 6.0: New Features & Updates (06/2016)
Here is a small selection of new or updated features released in BayesiaLab 6.0:
- A Function Node is an entirely new type of node, which provides an efficient and precise way of handling numerical functions within Bayesian Networks. Furthermore, Function Nodes allow you to build summaries of Joint Probability Distributions, which can be used for designing Reflexive Bayesian Networks.
- The Targeted Evaluation tools now feature numerous new metrics for evaluating the quality of networks with regard to specific modeling objectives, such as Classification, Forecasting, Prioritization, and Posterior Probability Estimation.
- The Multi-Target Evaluation tool offers a powerful way of measuring and comparing the performance of networks consisting of variables with different representations, e.g. different discretization or aggregation schemes.
- The Consistency metric can now also be used as a tool for evaluating the quality of networks.
- The Data Discretization Toolbox now offers a very comprehensive set of methods for univariate, bivariate, and multivariate discretizations.
- A Variable Selection algorithm is now available in conjunction with all Supervised Learning algorithms.
- The BayesiaLab WebSimulator, our new solution for sharing interactive models via the web with any audience, now offers Multi-Target Adaptive Questionnaires.