BayesiaLab 4.4

Category Intelligent Software>Bayesian Network Systems/Tools and Intelligent Software>Data Mining Systems/Tools

Abstract BayesiaLab, from Bayesia gives you a complete laboratory for manipulating Bayesian networks: develop your decision models through expertise and/or automatically from your data; quickly assimilate the represented knowledge using a set of original analytical tools; use the models in interactive mode or in batch, discover your optimal action policies by using reinforcement learning.

BayesiaLab Application Examples include:

1) Modeling and simulation of dynamic systems;

2) Global Risk Analysis and Security Policy;

3) Mining the Customer Database;

4) Fraud Detection;

5) Satisfaction questionnaire analysis;

6) Experience feedback exploitation;

7) Intrusion detection;

8) Text Mining;

9) DNA Microarray Analysis; and

10) Health Trajectory Analysis.

BayesiaLab features/capabilities include:

Develop your decision models through expertise --

1) Ergonomic node edition panel providing:

2) Constraint nodes to express constraint that hold between nodes.

3) Enhanced traceability and documentation thanks to hypertext comments associated to the graph, the nodes and the arcs.

4) Color node and arc tagging to semantically group your variables and probabilistic relations.

Automatic learning or updating of your models from your data (text files and databases) --

1) Learning conditional probabilities for a given network.

2) Discovering of all the probabilistic relations that hold in your data base (Association discovery).

3) Supervised learning entirely devoted to characterizing a target variable.

4) Selection of the minimal subset of variables correlated to the target variable.

5) Bayesian Clustering - to invent new concepts.

6) Robust Missing value processing.

7) Validation tools qualifying the obtained models (confusion matrix, lift and Roc curves).

Quickly assimilate the represented knowledge using a set of original analytical tools –

1) Strength of the probabilistic relations (arc's thickness and HTML report);

2) Amount of information brought to the target node/modality;

3) Type of probabilistic relations;

4) Complete HTML analysis report of the target variable;

5) HTML report of the evidence set analysis;

6) Causal analysis (essential graphs);

7) Automatic network lay outing algorithms.

Use the models in interactive or batch mode –

1) Positive, negative and soft evidences on the variable states;

2) Simulation of "What-if" scenarios with probability variation highlighting;

3) Adaptive questionnaires taking into account Costs and Information Gains;

4) Off-line Tagging of new cases contained in a file;

5) Robust imputation algorithm to complete data with missing values.

Introduce the temporal dimension into your models –

1) Compact representation of Dynamic Bayesian networks;

2) Time node for an explicit use of time in the equations;

3) Temporal simulation step by step or by period with a graphical view of the probability evolution;

4) Observations file to specify the context of the scenarios.

Representation, evaluation and learning of your action policies –

1) Decision nodes for modeling your actions;

2) Quality tables associated to Decision nodes for a direct representation of action policies;

3) Utility nodes to valuate the states and to associate cost/gains to node modalities;

4) Reinforcement learning algorithms to automatically discover policies that optimize the expected sum of utilities.

Complete interoperability --

1) Connection with your databases by using JDBC/ODBC;

2) SQL Interface;

3) Exportation of Bayesian networks, tables, equations, graphs, matrices and reports by using simple copy & paste, as images, numerical data and HTML texts.

Note: For the additional features/capabilities of this advanced product - (see G6G Abstract Number 20212R) - BayesiaLab 5.0.

System Requirements

Contact manufacturer.

Manufacturer

Manufacturer Web Site BayesiaLab 4.4

Price Contact manufacturer.

G6G Abstract Number 20212

G6G Manufacturer Number 100385