Our Software: A Complete Modeling Environment

Our software products offer a comprehensive modeling environment based on decision-theoretic methodologies.

An important design criterion for our products has been from the very start that they should allow for a complete modeling freedom. Models constructed by means of our software do not bend reality to what they can do but rather focus on doing whatever the problem at hand demands.  Because no exact algorithms exist for some types of models, our software is equipped with a suite of approximate stochastic sampling algorithms, capable of solving any models created by the users.

The list below describes our products and their scope.

GeNIe

Suitable for: anyone

Runs on: desktop operating systems (Windows, macOS, Linux)

GeNIe is an interactive model building, learning, and exploration tool. Its intuitive graphical interface makes it easy to construct and examine decision-theoretic models (Bayesian networks, dynamic Bayesian networks, influence diagrams, equation-based and hybrid systems).

Read more…

SMILE

Suitable for: programmers

Runs on: multiple operating systems and programming languages

SMILE is a programming library for probabilistic graphical models. SMILE can be embedded into existing user software, enhancing it with decision modeling capabilities. Our interactive software modules are all powered by SMILE. Low memory footprint and high performance ensure that SMILE is perfectly suited for software running in any environment, from data center through desktop to mobile and embedded.

SMILE’s native API is platform-independent C++ but we also provide wrappers for Python, R, Java, .NET and COM.

Read more…

BayesBox

Suitable for: any organization

Runs on: web

BayesBox allows for sharing probabilistic graphical models through an interactive web-based repository created for internal or public use.

BayesBox is available as a software installable and fully controlled by the user on-premises, or hosted as a service by BayesFusion. It is customizable to user needs. The user can upload any number of models organized into any number of categories. The website can be customized to reflect customer brand, including a name, a logo, and a color scheme. Optionally, access can be controlled through login/password.

Read more…

BayesMobile

Suitable for: anyone

Runs on: iPhone/iPad (Android coming soon)

BayesMobile allows for using Bayesian network models on mobile devices.

Read more…