BayesFusion, LLC, provides decision modeling software based on decision-theoretic principles. Our key product is GeNIe Modeler, a tool for modeling and learning with Bayesian networks, dynamic Bayesian networks, and influence diagrams. Our software library, SMILE Engine, allows for including our methodology in customers’ applications, which can be written in a variety of programming languages (e.g., C++, Python, Java, R, .NET). SMILE.COM allows for easy integration with MS Excel. You can visit our interactive Bayesian model repository to see SMILE in action.
We also offer training, scientific consulting, and custom software development.
The areas of application of our software include:
Decision Modeling
Decision-theoretic approach is about optimality. Anybody can make a decision but it is the detail that matters in the competitive world. Small differences between individuals and organizations can make the whole difference. This holds for such applications as diagnosis, prognosis, customer acquisition and retention and many others. Our software has been embedded in a variety of applications that need to make optimal choices efficiently.
Strategic Planning
What can an organization do to become more competitive? What are the opportunities for improving the working of an organization? Our software (GeNIe and QGeNIe) allows for modeling of organizations and calculating the optimal policies to achieve a desired objective. Causal discovery procedures allow for learning the structure of a system and finding what drives it. Once we know the drivers, we can find which of them are most effective in achieving the objectives.
Data Science
Mark Twain once wrote: “A man who does not read has no advantage over a man who cannot read.” A man who does not analyze his data has no advantage over a man who has no data. We offer modern data science methodologies to help our customers with their data analytics needs. While Bayesian methodologies are capable of capturing expert knowledge, having data makes the resulting models more accurate. With sufficient amount of data, models can be built automatically.