Configure the modules
Module configuration includes some preliminary module settings as well as their roles in the model.
Before starting to configure our flowchart, we still need to initiate some modifications.
1. Dealing with the features Selector
Having dismissed the data loader , let us deal with the features selector before splitting our dataset. Afterwards, we can finally start running our project.
As mentioned earlier, the iris flowers are classified according to the following distinctive relevant features:
petal width,
petal length,
sepal width,
sepal length.
Those are exactly the information we shall need to insert in our features selector.
Click on ‘Parameters’ then ‘Get columns from the dataset’. Under ‘Get columns’, the ‘selected columns’ instantly unlocks to reveal a scroll down menu of the iris classification features.
We need to include all features by adding them one by one.
To do so, click successively on every feature and follow up until the last one in the list. Click on the button ‘Add ’ or press the Enter key to validate each entry .
For selected label , however, as an output we shall need but ‘variety ’ since it is the only result we would like to be delivered. Also, enter ‘variety’ in the selected label category.
Once it is done, click on ‘Save’.
3. Splitting training and testing data with the labeled data splitter
The role of the labeled data splitter , as its name suggests, is to divide datasets into training and testing sets. Upon clicking on the module, a slider enables to variate this ratio at will.
N.B For machine learning purposes, the train proportion is usually greater than the test with an optimal value between 0.5-0.8
Last updated
