Netica Application is a comprehensive tool for working with Bayesian belief nets and decision nets (influence diagrams). It can build, learn, modify, transform and store nets, as well as answer queries or find optimal solutions using its powerful inference engine. This version has many new features and many more are currently under development.
Netica can also work on a Mac. More Info
Click here for new features.
• Compiles Bayes nets into junction trees of cliques for fast probabilistic reasoning.
• Can learn probabilistic relations from data (including EM and gradient descent learning).
• Generates presentation quality graphics which can be transferred to other documents, including SVG graphics.
• Allows the entry of probabilistic relations by equation, with an extensive built-in library of probabilistic functions and other mathematical functions. The equations can be deterministic or probabilistic, and for discrete or continuous variables.
• Provides easy graphical editing of Bayes nets and influence diagrams, including:
• Cutting and pasting of nodes and nets without losing their probabilistic relations.
• Many ways of displaying nodes (bar graphs, meters, etc.)
• Links with bends to keep complex diagrams orderly.
• Allows entering comments to document each node, keeps track of author, when changed, etc.
• Unlimited levels of undo/redo.
• Can create and work with sets of nodes including color-coding nodes.
• Comment windows for nodes, links and states.
• Dynamic scrolling and mouse wheel supported everywhere.
• Can find optimal decisions for sequential decision problems (i.e. later decisions are dependent on the results of earlier ones).
• Can test the performance of a Bayes net using a file of cases. Netica will print out a confusion matrix, error rate, logarithmic and quadratic (Brier) scoring rule results, calibration table and surprise indexes for each node desired.
• Can do utility-free sensitivity analysis.
• Can reverse individual links and “sum out” nodes of influence diagrams or belief nets, for model exploration and refinement.
• Supports disconnected links, which makes possible libraries of probabilistic relationships.
• Has facilities to enter and update individual cases, store them in their own files, and apply them to other Bayes nets.
• Has facilities for the easy discretization of continuous variables.
• Has no built-in limits on the size or complexity of nets, so they are limited only by available memory.
• Can work hand-in-hand with the Netica API Programmer’s Library.
• Can directly connect with a database or Excel spreadsheet for learning, reading cases, testing a net, etc.
• Has binary .neta format for faster and smaller files (text-based .dne files still have complete support).
• Can obfuscate nets, and can work with encrypted Bayes nets, to protect your intellectual property.
• Can automatically read in a case, compile the Bayes net when it is read from disk, and has auto-discretization, based on case files.