Feature Transform

Hardik Chheda Updated by Hardik Chheda

After selecting the required model, Tellius displays the model page with steps that you need to perform to create the Point-N-Click model.

Feature Transform is the first step in the process. Here you select the target feature that you want to transform and select other necessary options.

Now, specify appropriate values in the following fields:

Note: The fields on the page differs based on the type of model you select.
  • Target variable: Select the fields based on which you want to build the predictive analysis. The field contains only the continuous variables.
  • Handle null: Select the option how you want to handle the null values from the selected target variable. The field is visible only after selecting the target variable.
  • Dataset for training: Select the dataset for training your model.
    You can select one of the following option:
    • Complete dataset: The complete dataset from the selected business view is used for the model.
    • Sample dataset: The partial dataset from the selected business view is used for the model. For this option, you can specify the percentage of the data that you want to use.
    • Sample Percent (%): Specify the percentage of records that you wat to use for the model. The field is visible only when the Sample dataset option is selected.
  • Training set split ratio: Enter the percentage of records that you want to use to train the model vs. test the sample. The model picks the random samples based on the specified sampling ratio. Tellius runs the model on the remaining records as a test sample based on the learning.
    For example, if the sampling ratio is specified as 0.7 (70%). Tellius will train the model on random 70% of the dataset and use the remaining 30% of the dataset for testing the model.
  • NFolds: Enter the number of folds to be run for cross validation.
  • Evaluation Metric: Select the appropriate evaluation metric for your model.
  • Features: The table containing the data columns of dataset. Select the check boxes for the columns that you want to include in the analysis. Select check boxes for individual columns or select the check box in the table header to select all columns.

After specifying the values, click the Next button to complete the Model Selection process.

Did we help you?


Model Selection