68 строки
1.8 KiB
Markdown
68 строки
1.8 KiB
Markdown
|
---
|
||
|
|
layout: default
|
||
|
|
title: Description of Database Tables
|
||
|
|
---
|
||
|
|
|
||
|
|
## SQL Database Tables and Stored Procedures
|
||
|
|
-----------------------
|
||
|
|
|
||
|
|
Below are the data tables that you will find in either the `{{ site.db_name }}_R` or `{{ site.db_name }}_Py` databases after deployment.
|
||
|
|
|
||
|
|
<table class="table" >
|
||
|
|
<thead>
|
||
|
|
<tr>
|
||
|
|
<th>Table</th>
|
||
|
|
<th>Description</th>
|
||
|
|
</tr>
|
||
|
|
</thead>
|
||
|
|
<tbody>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.Label_Names</td>
|
||
|
|
<td>Names associated with each of the 20 labels</td>
|
||
|
|
</tr>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.Metrics</td>
|
||
|
|
<td>Metrics for the trained model</td>
|
||
|
|
</tr>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.Model</td>
|
||
|
|
<td>Trained model</td>
|
||
|
|
</tr>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.News_Test</td>
|
||
|
|
<td>News items in the test set</td>
|
||
|
|
</tr>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.News_To_Score</td>
|
||
|
|
<td>News items to score</td>
|
||
|
|
</tr>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.News_Train</td>
|
||
|
|
<td>News items for training the model</td>
|
||
|
|
</tr>
|
||
|
|
</tbody>
|
||
|
|
</table>
|
||
|
|
|
||
|
|
The following stored procedures are used in this solutions in either the `{{ site.db_name }}_R` or `{{ site.db_name }}_Py` databases:
|
||
|
|
<table class="table" >
|
||
|
|
<thead>
|
||
|
|
<tr>
|
||
|
|
<th>Stored Procedure</th>
|
||
|
|
<th>Description</th>
|
||
|
|
</tr>
|
||
|
|
</thead>
|
||
|
|
<tbody>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.evaluate</td><td>Evaluates the performance of the model</td>
|
||
|
|
</tr>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.initial_Run_Once_R <br/> dbo.initial_Run_Once_Py</td><td>Runs the training workflow natively in SQL for this solution</td>
|
||
|
|
</tr>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.score</td><td>Creates features on the fly for the testing set and makes predictions </td>
|
||
|
|
</tr>
|
||
|
|
<tr>
|
||
|
|
<td>dbo.train_model</td><td>Creates features on the fly for the training set and train the multiclass logistic regression model</td>
|
||
|
|
</tr>
|
||
|
|
</tbody>
|
||
|
|
</table>
|