You can think of Infer.NET as a programming language for building probabilistic models. Once you have a model, you can observe some values in your model and then query the model for unknown values - the answers are returned as distributions. This
model-based approach is very general and allows you to solve many types of problem including those in the areas you mention. However you do need to build a model for the problem you are trying to solve; Infer.NET does not provide any vertical solutions,
nor any automatic data preprocessing that you might need in such domains. The
Resources and References section of the user guide points to some background material for understanding the concepts behind Infer.NET.
John
Marked as answer bySuresh GorakalaSunday, October 7, 2012 4:41 AM
You can think of Infer.NET as a programming language for building probabilistic models. Once you have a model, you can observe some values in your model and then query the model for unknown values - the answers are returned as distributions. This
model-based approach is very general and allows you to solve many types of problem including those in the areas you mention. However you do need to build a model for the problem you are trying to solve; Infer.NET does not provide any vertical solutions,
nor any automatic data preprocessing that you might need in such domains. The
Resources and References section of the user guide points to some background material for understanding the concepts behind Infer.NET.
John
Marked as answer bySuresh GorakalaSunday, October 7, 2012 4:41 AM