class ui.module.MulticlassClassificationTask

Defines multiclass classification task.

Properties

MulticlassClassificationTask.num_classes

Number of classes

int

MulticlassClassificationTask.class_names

Names of the classes

list[str]

MulticlassClassificationTask.metric_to_monitor

Name of the default metric to monitor during validation.

str

MulticlassClassificationTask.metric_monitoring_mode

Whether to store values for maximum or minimum metric value.

MetricMonitoringMode

MulticlassClassificationTask.loss_function

The default loss function to use for training.

LossFunction

Methods

MulticlassClassificationTask.__init__(self, class_names)

from monad.ui.module import MulticlassClassificationTask

# define multiclass classification task with brand names
TARGETS = ["north_face", "adidas", "tommy_hilfiger", "hugo", "lacoste", "gap"]
task = MulticlassClassificationTask(class_names=TARGETS)

class_names: list[str]
The values of the target feature that will be used as class labels for multiclass classification.