The following node is available in the Open Source KNIME predictive analytics and data mining platform version 2.7.1. Discover over 1000 other nodes, as well as enterprise functionality at
http://knime.com.
X-Partitioner
This node is the first in a cross validation loop. At the end of the loop there must be a X-Aggregator to
collect the results from each iteration. All nodes in between these two node are executed as many times as
iterations should be performed.
Dialog Options
- Number of validations
-
The number of cross validation iterations that should be performed.
- Random sampling
-
If checked, the partitions are sampled randomly from the input table, otherwise it is cut into consecutive
pieces.
- Stratified sampling
-
If checked, the partitions are sampled randomly but the class distribution from the column selected
below is maintained.
- Random seed
-
For random and stratified sampling you can choose a seed for the random number
generator in order to get reproducible results. Otherwise you get different
partitions every time.
- Class column name
- The name of the column with the class labels.
- Leave-one-out
-
Performs a leave-one-out cross validation, i.e. there are as many iterations as data points and in each
iteration another point's target value is predicted by using all remaining points as training set.
Ports
Input Ports
0 |
The datatable that is to be split |
Output Ports
0 |
The data table with the training data |
1 |
The data table with the test data |
This node is contained in KNIME Base Nodes
provided by KNIME GmbH, Konstanz, Germany.