Mli_surrogates configuration

mli_lime_method

LIME method. (String)

Default value 'k-LIME'

Choose LIME method to be used for creation of surrogate models.

mli_use_raw_features

Use original features for surrogate models. (Boolean)

Default value True

Choose whether surrogate models should be built for original or transformed features.

mli_ts_use_raw_features

Use original features for time series based surrogate models. (Boolean)

Default value False

Choose whether time series based surrogate models should be built for original features.

mli_vars_to_pdp

Number of features for Surrogate Partial Dependence Plot. Set to -1 to use all features. (Number)

Default value 10

Set maximum number of features for which to build Surrogate Partial Dependence Plot. Use -1 to calculate Surrogate Partial Dependence Plot for all features.

mli_nfolds

Cross-validation folds for surrogate models. (Number)

Default value 3

Set the number of cross-validation folds for surrogate models.

mli_qbin_count

Number of columns to bin for surrogate models. (Number)

Default value 0

Set the number of columns to bin in case of quantile binning.

mli_sample_size

Sample size for surrogate models. (Number)

Default value 100000

The sample size, number of rows, used for MLI surrogate models.

mli_num_quantiles

Number of bins for quantile binning. (Number)

Default value 10

Number of bins for quantile binning.

mli_drf_num_trees

Number of trees for Random Forest surrogate model. (Number)

Default value 100

Number of trees for Random Forest surrogate model.

mli_drf_max_depth

Max depth for Random Forest surrogate model. (Number)

Default value 20

Max depth for Random Forest surrogate model.

klime_lambda

Regularization strength for k-LIME GLM’s. (List)

Default value [1e-06, 1e-08]

Regularization strength for k-LIME GLM’s.

klime_alpha

Regularization distribution between L1 and L2 for k-LIME GLM’s. (Float)

Default value 0.0

Regularization distribution between L1 and L2 for k-LIME GLM’s.

mli_max_numeric_enum_cardinality

Max cardinality for numeric variables in surrogate models to be considered categorical. (Number)

Default value 25

Max cardinality for numeric variables in surrogate models to be considered categorical.

mli_max_number_cluster_vars

Maximum number of features allowed for k-LIME k-means clustering. (Number)

Default value 6

Maximum number of features allowed for k-LIME k-means clustering.

use_all_columns_klime_kmeans

Use all columns for k-LIME k-means clustering (this will override `mli_max_number_cluster_vars` if set to `True`). (Boolean)

Default value False

Use all columns for k-LIME k-means clustering (this will override mli_max_number_cluster_vars if set to True).