BayesiaLabは、ベイジアンネットワーク技術を利用するためのツールキットです。BayesiaLabにより、ベイジアンネットワークの強力なパワーを容易に利用することが可能になります。
BayesiaLabは、手作業によるベイジアンネットワークの作成/編集機能、データからのベイジアンネットワークの自動抽出機能を持っています。データマイニングの強力なツールとしてもご利用いただけます。より詳細な機能については、下記の表を参照してください。
BayesiaLabには、スタンダード版とプロフェッショナル版の2種類があります。下記の表は、BeyesiaLabの機能のスタンダード版とプロフェッショナル版の比較表でもあります。
| Functions |
Standard Edition |
Professional Edition |
| Inference |
Exact Inference with Junction Tree |
X |
X |
| Approximate Inference with Importance Sampling |
X |
X |
| Interactive inference based on a file of observations |
X |
X |
| Interactive Bayesian updating based on a file of observations |
X |
X |
| Adaptive Questionnaire with respect to a target variable |
X |
X |
| Adaptive Questionnaire with respect to a target modality |
X |
X |
| Batch Labeling of the target variable |
- |
X |
| Batch Inference of the Not Observable variables |
- |
X |
| Batch Joint Probability |
- |
X |
| Markov Blanket exportation (Available on subscription only) |
SAS Macro |
- |
X |
| Data |
Data generation with MCMC |
- |
X |
| JDBC/ODBC connection |
- |
X |
| Database saving |
- |
X |
| Imputation of the Missing Values |
- |
X |
| Dictionaries |
Comments |
- |
X |
| Colour categories |
- |
X |
| Classes |
- |
X |
| Observation cost |
- |
X |
| Temporal indexes |
- |
X |
| Modality values |
- |
X |
| Modality names |
- |
X |
| Exportation of the dictionaries |
- |
X |
| Discretization of continuous variables |
Manual based on the repartition function |
X |
X |
| Equal distances |
X |
X |
| Equal frequencies |
X |
X |
| Decision tree |
- |
X |
| Aggregation of discrete modalities |
Manual |
X |
X |
| Manual based on the correlation with a target variable |
X |
X |
| Semi-Automatic wrt the correlation with a target modality |
X |
X |
| Decision tree based on the correlation with a target modality |
- |
X |
| Missing values processing |
Filtering |
X |
X |
| Replacement |
X |
X |
| Inference |
X |
X |
| Association discovery |
EQ |
- |
X |
| SopLEQ |
- |
X |
| Taboo Search |
X |
X |
| Taboo Order |
- |
X |
| Target node Characterization |
Naive |
X |
X |
| Augmented Naive |
X |
X |
| Sons & Spouses |
- |
X |
| Markov Blanket |
- |
X |
| Augmented Markov Blanket |
- |
X |
| Minimal Augmented Markov Blanket |
- |
X |
| Semi-Supervised Learning |
- |
X |
| Clustering |
Variable clustering |
- |
X |
| Data clustering |
- |
X |
| Multiple clustering |
- |
X |
| Targeted Evaluation |
Multiple thresholds |
- |
X |
| Global Precision |
X |
X |
| Confusion Matrix |
X |
X |
| Lift chart |
- |
X |
| Gain chart |
- |
X |
| ROC curve |
- |
X |
| Automatic layout algorithms |
Symmetric |
X |
X |
| Dynamic |
X |
X |
| Genetic |
- |
X |
| Mutual Information |
- |
X |
| Random |
X |
X |
| Graphical Network analysis |
Global strength of the Arcs |
X |
X |
| Pearson's Correlation |
X |
X |
| Target node |
X |
X |
| Target modality |
X |
X |
| Influence paths |
X |
X |
| Causal analysis (essential graphs) |
X |
X |
| HTML Analysis Report |
Arcs |
X |
X |
| Target Node |
X |
X |
| Evidence set (contradiction analysis) |
X |
X |
| Special nodes |
Hidden |
X |
X |
| Decision |
X |
X |
| Utility |
X |
X |
| Constraint |
X |
X |
| Dynamic Bayesian networks |
X |
X |
| Action policy learning |
Static Bayesian networks |
X |
X |
| Dynamic Bayesian networks |
X |
X |
| Graphics |
Histogram |
X |
X |
| Repartition function |
X |
X |
| Occurrence matrix with Khi2 test |
X |
X |
| 2D Scatter points |
- |
X |
| 3D Scatter points |
- |
X |
| Bubble chart |
- |
X |
| Multilingual |
English |
X |
X |
| French |
X |
X |
| Japanese |
X |
X |
| Cross-Platform (Java technology) |
X |
X |