To begin, you need to find a dataset relevant to your problem. You can search for datasets from the Distil Home page or the Select Data view.


In the Select Data view, you can search for datasets using keywords (e.g. vehicle acceleration) or by expressing your intent in natural language (e.g. Predict vehicle acceleration).

Select Data
Click to enlarge

Search for a dataset

To find datasets related to automotive attributes:
  1. Enter acceleration in the Search field on the Home page or the Select Data view. As you type, Distil returns each dataset that contains the keyword in its list of features or full description.
  2. Review the d_196_ag_dataset_TRAIN result to determine its relevancy.
    • The Top Features, which indicate the most important features (columns), list acceleration and several other quantitative features such as displacement and weight.
    • The Description, inferred from text in the dataset, indicates that the dataset is about engine configurations.
  3. Click More Details to see a complete summary.
    Expand to view more dataset details
  4. Click Name: d_196_ag_dataset_TRAIN to choose the automotive dataset for further inspection.
    Choose a dataset to begin selecting which features to include in your model

Next steps

Choose a feature in the dataset that you want to predict with your model.