BayesiaLab 5.0

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 5.0 -- New Features:

BayesiaLab Data --

1) New format for arc constraint dictionary;

2) Disambiguation of the names of classes, nodes and states in the dictionaries;

3) Byte Order Mask (BOM) taken into account when exporting dictionaries;

4) State virtual number dictionary - Dictionaries of state virtual numbers can be imported or exported;

5) Local structural coefficient dictionary - Dictionaries of local structural coefficients can be imported or exported;

6) Rearranging dictionaries - The menus used to import and export dictionaries has been rearranged;

7) Database report - A database report can be generated via the contextual menu;

8) Choice of unnecessary columns when associating the database;

9) Row identifier - A new type of column is available when a database is imported or associated: setting a column of the database as a row identifier allows defining an identifier for each row; and

10) Information about the other types of columns.

BayesiaLab Network --

1) Costs greater or equal to 1 - Now, all the costs associated to the nodes must be greater or equal to 1. If Not, the corresponding nodes are considered as Not observable;

2) Select nodes with missing values - allows selecting the nodes having a percentage of missing values greater than a given threshold; and

3) Select nodes with assessments - allows selecting the nodes having assessments.

BayesiaLab Learning --

1) K-Means Clustering - allows clustering data in an unsupervised way thanks to the k-means clustering algorithm, in order to find partitions of homogeneous elements;

2) New parameters for Taboo learning;

3) New parameters for Taboo Order learning;

4) Structure equivalent example number for each learning algorithm; and

5) Keeping learning parameters - For each parameterized learning algorithm, the chosen parameters are saved in the network and can be reused later in various tools.

BayesiaLab Inference --

1) Observed mean/value - Now, a target mean/value can be set as an evidence for a node;

2) Fixing current probabilities;

3) Relaxing probabilities;

4) Causal inference - In addition to the standard exact inference, it is possible to make causal inference, i.e. to consider a set of nodes as causal nodes in the Bayesian network;

5) New database save dialogs; and

6) Selecting database rows with an identifier.

BayesiaLab Analysis --

1) A variation editor for analysis;

2) Upgrade of the target dynamic profile;

3) Target optimization tree - This function allows searching for several policies in order to optimize the target node;

4) Target interpretation tree - This function allows generating a tree that helps interpreting the target node;

5) Kolmogorov-Smirnov (K-S) test in network global performance - The K-S test has been added to the network global performance when the network has learning and test databases;

6) Extract database in global performance - allows extracting the data on which the computed log-likelihood is inside a given interval;

7) Fix reference in influence analysis - allows keeping the value of the a priori probability (red line) as a reference if new evidence is set on the monitors;

8) Arc's mutual information - mutual information can be computed for each arc of the network;

9) Filtered values are taken into account in mutual information;

10) Colored borders in influence analysis;

11) Translucent nodes skipped in analyses;

12) Filtered states skipped in the target report;

13) Translucent arcs skipped in relationship analysis;

14) Hidden variable discovery - This report allows computing the G-test and the independence probability between two variables of the network bounded by a path of length one or two and which is Not a V-structure;

15) Target optimization - The target state optimization has been transformed to become the target optimization. It is now possible to optimize a state of a node or the mean of a node; and

16) Probability analysis report for the target state - This analysis report is available for any target node.

BayesiaLab Monitors --

1) New menu Monitor;

2) Log-likelihood is displayed in the monitor information panel;

3) Sorting monitors -

4) Fitting the monitor's size -

5) Scientific notation - displays the probabilities in scientific notation in the monitors and in the monitors' tooltips;

6) Tooltip with extended values for the information panel of the monitors - Values in the information panel is displayed in the tooltip with a greater number of decimals;

7) Fix and relax probabilities; and

8) Display properties saved - The various display properties of the monitors are now saved with the network.

BayesiaLab Interface --

1) Three (3) state stop button;

2) Colored names in reports;

3) Colored line plot;

4) Layer intersection;

5) Import/Export in property editors - It is now possible to import and export dictionaries of properties in the editors of costs, constants, temporal indices, local structural coefficients and state virtual numbers; and

6) Sortable tables in property editors.

BayesiaLab Knowledge Elicitation --

1) Assessments - The mechanism of knowledge elicitation of BayesiaLab allows a set of experts to provide assessments on the conditional probability tables of nodes. A “facilitator” can mediate to ask relevant questions to experts and get their assessments;

2) Experts - The assessment mechanism defines experts associated with the network;

3) Online assessment editing;

4) Assessment tools - a) provides access to various tools for exporting assessments and b) provides access to the assessment sensitivity analysis;

5) Assessment sensitivity analysis - This tool allows, through sampling, to visualize graphically the impact of the different assessments of a network’s variables on each state of target variables; and

6) Assessment report.

BayesiaLab Tools --

1) Exporting assessments;

2) Shuffle on data in cross-validation - In cross-validation tools, data are shuffled before slicing the database in order to avoid biases due to ordered data;

3) Joint comparison - This tool is used to compare the joint probability distributions from two (2) different files;

4) Structural coefficient modified in cross-validation - In cross-validation tools, the structural coefficient of the reference network is recomputed for the generated networks by multiplying it with a final learning weight divided by the initial learning weight;

5) Database stratification in cross-validation - In cross-validation tools, if the initial database is stratified, the generated databases will be also stratified with the same parameters;

6) Structural coefficient analysis;

7) Data perturbation - allows avoiding the local minima of the network’s structure in which the network can be trapped; and

8) Multi-quadrant analysis - This tool allows analyzing a network having a variable called “selector” which must be different from the target node. For each state of this variable, one of the three (3) available analysis will be performed on the variables of the network with respect to the target.

BayesiaLab Help --

1) Analysis of the use of BayesiaLab.

BayesiaLab Miscellaneous --

1) Analyst edition - In addition to the standard, professional and educational editions, a new edition named Analyst is now available. The Analyst edition is only oriented towards network analysis. It does Not contain data-mining tools.

Note: For the additional features/capabilities of this advanced product - (see G6G Abstract Number 20212) - BayesiaLab 4.4.

Manufacturer

Manufacturer Web Site BayesiaLab 5.0

Price Contact manufacturer.

G6G Abstract Number 20212R

G6G Manufacturer Number 100385