Data
Introduction
Bayes Server allows access to data sources such as databases or spreadsheets, so that data can be managed in a standard way, but easily accessed throughout Bayes Server.
For example:
- Set evidence on a Bayesian network or Dynamic Bayesian network through Data Explorer.
- Run a batch of queries on a Bayesian network or Dynamic Bayesian network through. See Batch query for more information.
- Learn the parameters of a Bayesian network or Dynamic Bayesian using a data source.
Data
Standard data, as opposed to DBN data, is data that can be mapped to non DBN (standard) Variables in a Bayesian network.
This can be thought of as a table in a database, where each column refers to a Variable, and each row represents a separate unit of information, such as a Customer.
Example data:
| Case | Country | Age |
|---|---|---|
| 1 | US | 35 |
| 2 | UK | 41 |
| 3 | Japan | 25 |
| 4 | US | 62 |
DBN Data
DBN Data is data that contains one or more time series/sequences, and can be mapped to DBN Variables in a Dynamic Bayesian network.
DBN data must be stored in a single table, where each Time Series/sequence is identified by a Case Id Column.
For an individual time series/sequence, each row must be labeled with a time value. For more information about the time value see Time Column.
Example DBN data:
| Case | Time | Transition | Obs1 | Obs2 |
|---|---|---|---|---|
| 1 | 0 | Cluster 0 | 12.4 | 15.5 |
| 1 | 1 | Cluster 1 | 14.2 | 13.45 |
| 2 | 0 | Cluster 1 | 9.3 | 8.6 |
| 2 | 1 | Cluster 1 | 12.3 | 14.0 |
| 2 | 2 | Cluster 1 | 18.2 | 13.5 |
Case
The term Case is used to refer to an individual row of non DBN data, coupled with any DBN data that corresponds to that row.