README.md

metadata

task_categories:
  - image-classification

AutoTrain Dataset for project: cancer-lakera

Dataset Description

This dataset has been automatically processed by AutoTrain for project cancer-lakera.

Languages

The BCP-47 code for the dataset's language is unk.

Dataset Structure

Data Instances

A sample from this dataset looks as follows:

[
  {
    "image": "<600x450 RGB PIL image>",
    "feat_image_id": "ISIC_0024329",
    "feat_lesion_id": "HAM_0002954",
    "target": 0,
    "feat_dx_type": "histo",
    "feat_age": 75.0,
    "feat_sex": "female",
    "feat_localization": "lower extremity"
  },
  {
    "image": "<600x450 RGB PIL image>",
    "feat_image_id": "ISIC_0024372",
    "feat_lesion_id": "HAM_0005389",
    "target": 0,
    "feat_dx_type": "histo",
    "feat_age": 70.0,
    "feat_sex": "male",
    "feat_localization": "lower extremity"
  }
]

Dataset Fields

The dataset has the following fields (also called "features"):

{
  "image": "Image(decode=True, id=None)",
  "feat_image_id": "Value(dtype='string', id=None)",
  "feat_lesion_id": "Value(dtype='string', id=None)",
  "target": "ClassLabel(names=['actinic_keratoses', 'basal_cell_carcinoma', 'benign_keratosis-like_lesions'], id=None)",
  "feat_dx_type": "Value(dtype='string', id=None)",
  "feat_age": "Value(dtype='float64', id=None)",
  "feat_sex": "Value(dtype='string', id=None)",
  "feat_localization": "Value(dtype='string', id=None)"
}

Dataset Splits

This dataset is split into a train and validation split. The split sizes are as follow:

Split name	Num samples
train	1200
valid	150