Unsupervised Metrics

This section documents the unsupervised metrics available in pytorch_tabnet.metrics.

Unsupervised Loss

Used for unsupervised pretraining, typically measures reconstruction error.

Example:

import torch
from pytorch_tabnet.metrics import UnsupervisedLoss
y_pred = torch.randn(3, 5)
embedded_x = torch.randn(3, 5)
obf_vars = torch.randint(0, 2, (3, 5)).float()
loss = UnsupervisedLoss(y_pred, embedded_x, obf_vars)
print(loss)

Unsupervised Metrics

Additional metrics for unsupervised learning tasks.

Example:

import torch
from pytorch_tabnet.metrics import UnsupervisedMetric
metric = UnsupervisedMetric()
y_pred = torch.randn(3, 5)
embedded_x = torch.randn(3, 5)
obf_vars = torch.randint(0, 2, (3, 5)).float()
score = metric(y_pred, embedded_x, obf_vars)
print(score)

Base Metrics

Base Metrics

Contains base classes and utilities for defining custom metrics.

Example:

import torch
from pytorch_tabnet.metrics import Metric
# See source for custom metric implementation