Classification: objectives and metrics

Name Used for optimization User-defined parameters Formula and/or description
Logloss +

use_weights

Default: true

Calculation principles

CrossEntropy +

use_weights

Default: true

Calculation principles

Precision

use_weights

Default: true

Calculation principles

Recall

use_weights

Default: true

Calculation principles

F1

use_weights

Default: true

Calculation principles

BalancedAccuracy

use_weights

Default: true

Calculation principles

BalancedErrorRate

use_weights

Default: true

Calculation principles

MCC

use_weights

Default: true

Calculation principles

Accuracy

use_weights

Default: true

Calculation principles

CtrFactor

use_weights

Default: true

Calculation principles

AUC*
  • use_weights

    Default: false

  • type

    Default: Classic for models with Logloss and CrossEntropy loss functions and Ranking for models with ranking loss functions.

Classic
The sum is calculated on all pairs of objects such that:

Refer to the Wikipedia article for details.

If the target type is not binary, then every object with target value and weight is replaced with two objects for the metric calculation:

  • with weight and target value 1
  • with weight and target value 0.

Target values must be in the range [0; 1].

Ranking

The sum is calculated on all pairs of objects such that:
NormalizedGini

use_weights

Default: true

See AUC.

Calculation principles

BrierScore

Calculation principles

HingeLoss

use_weights

Default: true

Calculation principles

HammingLoss

use_weights

Default: true

Calculation principles

ZeroOneLoss

use_weights

Default: true

Calculation principles

Kappa

Calculation principles

WKappa

Calculation principles

LogLikelihoodOfPrediction

Calculation principles

Name Used for optimization User-defined parameters Formula and/or description
Logloss +

use_weights

Default: true

Calculation principles

CrossEntropy +

use_weights

Default: true

Calculation principles

Precision

use_weights

Default: true

Calculation principles

Recall

use_weights

Default: true

Calculation principles

F1

use_weights

Default: true

Calculation principles

BalancedAccuracy

use_weights

Default: true

Calculation principles

BalancedErrorRate

use_weights

Default: true

Calculation principles

MCC

use_weights

Default: true

Calculation principles

Accuracy

use_weights

Default: true

Calculation principles

CtrFactor

use_weights

Default: true

Calculation principles

AUC*
  • use_weights

    Default: false

  • type

    Default: Classic for models with Logloss and CrossEntropy loss functions and Ranking for models with ranking loss functions.

Classic
The sum is calculated on all pairs of objects such that:

Refer to the Wikipedia article for details.

If the target type is not binary, then every object with target value and weight is replaced with two objects for the metric calculation:

  • with weight and target value 1
  • with weight and target value 0.

Target values must be in the range [0; 1].

Ranking

The sum is calculated on all pairs of objects such that:
NormalizedGini

use_weights

Default: true

See AUC.

Calculation principles

BrierScore

Calculation principles

HingeLoss

use_weights

Default: true

Calculation principles

HammingLoss

use_weights

Default: true

Calculation principles

ZeroOneLoss

use_weights

Default: true

Calculation principles

Kappa

Calculation principles

WKappa

Calculation principles

LogLikelihoodOfPrediction

Calculation principles