Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -1208,7 +1208,7 @@ The dataset presents eleven sections (one per regression task), with varying deg
|
|
| 1208 |
|
| 1209 |
## Partitions and versions
|
| 1210 |
|
| 1211 |
-
The sections are already partitioned into a predefined train-validation-test split for training assessors. Assessors need a particular kind of partitioning (mainly stratified by instance identifier to avoid contamination), so that's why the
|
| 1212 |
|
| 1213 |
The **main** branch contains the unaltered datasets, keeping the original values of the task and model characteristics, whereas the **normalised** branch contains the datasets properly normalised.
|
| 1214 |
|
|
|
|
| 1208 |
|
| 1209 |
## Partitions and versions
|
| 1210 |
|
| 1211 |
+
The sections are already partitioned into a predefined train-validation-test split for training assessors. Assessors need a particular kind of partitioning (mainly stratified by instance identifier to avoid contamination), so that's why the subsets are given.
|
| 1212 |
|
| 1213 |
The **main** branch contains the unaltered datasets, keeping the original values of the task and model characteristics, whereas the **normalised** branch contains the datasets properly normalised.
|
| 1214 |
|