Multiclasseer

multiclasseer

Metric MulticlassEER from torchmetrics (torchmetrics.classification.MulticlassEER)

When to invoke this skill

The user has predictions + ground truth and asks to evaluate with MulticlassEER, or mentions torchmetrics.classification.MulticlassEER directly, or wants the standard torchmetrics implementation.

Reference signature

from torchmetrics.classification import MulticlassEER

# MulticlassEER(num_classes: int, thresholds: Union[int, list[float], torch.Tensor, NoneType] = None, average: Optional[Literal['micro', 'macro']] = None, ignore_index: Optional[int] = None, validate_args: bool = True, **kwargs: Any) -> None

Library docstring

Compute Equal Error Rate (EER) for multiclass classification task.

.. math::
    \text{EER} = \frac{\text{FAR} + (1 - \text{FRR})}{2}, \text{where} \min_t abs(FAR_t-FRR_t)

The Equal Error Rate (EER) is the point where the False Positive Rate (FPR) and True Positive Rate (TPR) are
equal, or in practise minimized. A lower EER value signifies higher system accuracy.

As input to ``forward`` and ``update`` the metric accepts the following input:

    - ``preds`` (:class:`~torch.Tensor`): A float tensor of shape ``(N, C, ...)`` containing probabilities or logits
      for each observation. If preds has values outside [0,1] range we consider the input to be logits and will auto
      apply softmax per sample.
    - ``target`` (:class:`~torch.Tensor`): An int tensor of shape ``(N, ...)`` containing ground truth labels, and
      therefore only contain values in the [0, n_classes-1] range (except if `ignore_index` is specified).

As output to ``forward`` and ``compute`` the metric returns the following output:

    - ``mc_eer`` (:class:`~torch.Tensor`): If `average=None` then a 1d tensor of shape (n_classes, ) will
      be returned with eer score per class. If `average="macro"|"micro"` then a single scalar will be returned.

Additional dimension ``...`` will be flattened into the batch dimension.

The implementation both supports calculating the metric in a non-binned but accurate version and a
binned version that is less accurate but more memory efficient. Setting the `thresholds` argument to `None` will
activate the non-binned version that uses memory of size :math:`\mathcal{O}(n_{samples})` whereas setting the
`thresholds` argument to either an integer, list or a 1d tensor will use a binned version that uses memory of
size :math:`\mathcal{O}(n_{thresholds} \times n_{classes})` (constant memory).

Args:
    num_classes: Integer specifying the number of classes
    thresholds: Can be one of:

        - If set to `None`, will use a non-binned approach where thresholds are dynamically calculated from
          all the data. Most accurate but also most memory consuming approach.
        - If set to an `int` (larger than 1), will use that number of thresholds linearly s

Quick recipe

import torchmetrics.classification as _m
score = _m.MulticlassEER(y_true, y_pred)

Don'ts

• Don't reimplement when the library version handles edge cases (NaN, ties, empty inputs) better than a hand-rolled formula.
• Always check the library version's argument order — sklearn is (y_true, y_pred) while torchmetrics is (preds, target).