Resampler (DP_RES)

Resample input datasets using specified resampling techniques.

Resampling helps in providing more flexibility when handling imbalanced datasets.

Supported Resampling Techniques:

  1. RandomOverSampler:

  2. SMOTE

  3. RandomUnderSampler

  4. TomekLinks

  5. SMOTETomek

Args:

xy_train (bool): Flag indicating whether to resample the training dataset.

xy_test (bool): Flag indicating whether to resample the testing dataset.

xy (bool): Flag indicating whether to resample main x and y datasets.

resampler (str): The name of the resampling technique to use.

details (object): Object containing dataset details.

load_name (object): Name of a saved resampler object, for loading.

save_name (str): Name to save the resampler object as.

Sample Request

The request performs resampling operation using the SMOTE technique

{
    "block_id": 000,
    "project_id": 000,
    "parent_id": 000,
    "function_code": "DP_RES",
    "args": {
        "xy": true,
        "xy_train": false,
        "xy_test": false,
        "resampler": "SMOTE",
        "save_name": "test"
    }
}

Parameter Details

POST https://autogon.ai/api/v1/engine/start

Request Body

NameTypeDescription

project_id*

int

current project ID

block_id*

int

current block ID

parent_id*

int

parent block ID

function_code*

String

block's function code

args*

object

block arguments

Last updated