Charles River Analytics, developer of intelligent system solutions, announces the release of Figaro 4.0, a probabilistic programming language that helps people effectively reason, learn, and make decisions in the face of uncertainty. Figaro 4.0 includes many new features, including a new technique called Structured Factored Inference (SFI), an alternative means of reasoning on a model. SFI decomposes a model into smaller parts, which can then be solved by independently and combined together to produce an answer. SFI can significantly improve inference time since each portion of the model can be solved by the inference algorithm that is most efficient on that specific part of the model.
Other new features in Figaro 4.0 include:
- New algorithms, including:
- Gibbs sampling on factor graphs
- Parallel Importance sampling and particle filtering
- Improvements to existing algorithms
- New tools for non-deterministic testing
- Added examples from the book, Practical Probabilistic Programming
Figaro supports the development of rich probabilistic models and provides reasoning algorithms to draw useful conclusions from the available “noisy” evidence.
“Reasoning under uncertainty requires taking what you know and inferring what you don’t know,” explained Dr. Avi Pfeffer, Principal Scientist at Charles River. “Probabilistic reasoning is a well-established approach for reasoning under uncertainty. Typically, we create a probabilistic model over all the variables we’re interested in, observe the values of some of these variables, and query the model for values of other variables of interest. A number and variety of approaches exist to handle this, and new approaches are being developed. Figaro is designed help build and reason with this wide range of probabilistic models.”