Skip to main content

Getting started

Introduction

Bayes Server is a tool for modeling Bayesian networks, Causal models, Dynamic Bayesian networks and Decision graphs.

Bayesian networks are widely used in the fields of Artificial Intelligence, Causal AI, Machine Learning, and Data Science.

Some examples of how Bayesian networks are used are given below:

Bayesian networks

Bayesian networks are a graphical approach to modeling, using probability. In a network, nodes are used to represent variables, and links to indicate that one node influences another. This allows the relationship between variables to be visualized easily. Each node in a Bayesian network requires a probability distribution to be specified (conditional on its parents), and Bayes Server uses advanced algorithms to combine these distributions, in order to answer queries (questions/predictions).

More information about Bayesian networks.

Build

The steps to build a Bayesian network are as follows:

  1. Add Nodes
  2. Add Links
  3. Define Node Distributions

In Bayes Server, each of these steps can be performed manually of learned via data, or a mixture of both.

Add Nodes

  1. Manually add a node
  2. Add nodes from data
  1. Manually add links
  2. Structural learning

Define Node Distributions

  1. Manually define a distribution
  2. Parameter learning
info

If required, you can learn some nodes from data, and define some manually.

Evidence

Once you have built a network, but before you query it, you can set evidence on nodes in a number of ways:

  1. Manually via Network Viewer, clicking nodes
  2. Data Explorer
  3. Causal interventions

Query

Once you have built a network, you can then query it to perform reasoning, diagnostics, causal inference, prediction and much more.

  1. Query directly in Network Viewer
  2. Batch query
  3. Custom query
  4. Mesh query
  5. Causal query

Analyze

Once a network has been built, in addition to querying the network, you can perform advanced analysis, such as...

  1. Evidence Optimization
  2. Value of Information
  3. Log-Likelihood Analysis
  4. Automated Insight
  5. D-Separation
  6. Retracted Analysis
  7. Pattern Analysis
  8. Impact Analysis
  9. ... and more ...

Dynamic Bayesian networks

Dynamic Bayesian networks add the concept of time, to allow for time series/sequence modeling.

More information about dynamic Bayesian networks.

Decision graphs

Decision graphs add the concept of utilities (e.g. profits & costs) and decisions, to allow for decision making under uncertainty.

More information about Decision graphs.

info

Evidence Optimization is an alternative technique that can be used for decision making.