Learning From Cases Using Excel
1. Create an Excel Spreadsheet
2. Add Nodes to the Net
3. Discretize or Combine States
4. Add Link Structure
5. Learn CPTs
6. Use the Resulting Bayes Net
The next step is to have Netica learn the conditional probability tables, or = 4 && typeof(BSPSPopupOnMouseOver) == 'function') BSPSPopupOnMouseOver(event);" class="BSSCPopup" onclick="BSSCPopup('X_PU_CPT.htm');return false;">CPTs.
Choose Cases → Learn→ Incorp Case File from the menu, and from the open-file dialog box presented, choose the Excel file with the spreadsheet. When asked, enter a “degree” of 1.0 unless you are learning from more than one data source, and you want to weight each one. Netica will process the Excel spreadsheet, and fill each node’s conditional probability table.
Note: When learning from data sets, the top row does not necessarily need to have the node name (Netica can still learn the data if there isn't a row heading/node name). It is preferable to have a node name, but Netica can learn it without the name. If it does have a node name, it must exactly match and must not contain spaces or special characters. See case file format for more info.
If the = 4 && typeof(BSPSPopupOnMouseOver) == 'function') BSPSPopupOnMouseOver(event);" class="BSSCPopup" onclick="BSSCPopup('X_PU_Messages_window.htm');return false;">Messages window is visible, you will be able to see the progress as the percent of cases processed so far. You can abort the operation by pressing ctrl while holding down the mouse button for a while.