Skip to content

Nested cross-validation for unbiased predictions. Can be used with Scikit-Learn, XGBoost, Keras and LightGBM, or any other estimator that implements the scikit-learn interface.

Switch branches/tags

Latest commit

outer_kfolds and inner_kfolds renamed to outer_cv and inner_cv

Git stats


Failed to load latest commit information.


This repository implements a general nested cross-validation function. Ready to use with ANY estimator that implements the Scikit-Learn estimator interface.

Installing the pacakge:

You can find the package on pypi* and install it via pip by using the following command:

pip install nested-cv

You can also install it from the wheel file on the Releases page.

* we gradually push updates, pull this master from github if you want the absolute latest changes.


Be mindful of the options that are available for NestedCV. Some cross-validation options are defined in a dictionary cv_options. This package is optimized for any estimator that implements a scikit-learn wrapper, e.g. XGBoost, LightGBM, KerasRegressor, KerasClassifier etc.

-->See notebook for more examples


Here is a single example using Random Forest. Check out the example notebook for more.

from nested_cv import NestedCV
from sklearn.ensemble import RandomForestRegressor

# Define a parameters grid
param_grid = {
     'max_depth': [3, None],
     'n_estimators': [10]

NCV = NestedCV(model=RandomForestRegressor(), params_grid=param_grid,
               outer_cv=5, inner_cv=5, n_jobs = -1,

NestedCV Parameters

Name type description
model estimator The estimator implements scikit-learn estimator interface.
params_grid dictionary "dict" The dict contains hyperparameters for model.
outer_cv int or cv splitter class Outer splitting strategy. If int, KFold is default.
inner_cv int or cv splitter class Inner splitting strategy. If int, KFold is default.
cv_options dictionary "dict" Next section
n_jobs int Number of jobs for joblib to run (multiprocessing)

cv_options value options

metric : Callable from sklearn.metrics, default = mean_squared_error

      A scoring metric used to score each model

metric_score_indicator_lower : boolean, default = True

      Choose whether lower score is better for the metric calculation or higher score is better.

sqrt_of_score : boolean, default = False

      Whether or not if the square root should be taken of score

randomized_search : boolean, default = True

      Whether to use gridsearch or randomizedsearch from sklearn

randomized_search_iter : int, default = 10

      Number of iterations for randomized search

recursive_feature_elimination : boolean, default = False

      Whether to do feature elimination

predict_proba : boolean, default = False

      If true, predict probabilities instead for a class, instead of predicting a class

multiclass_average : string, default = 'binary'

      For some classification metrics with a multiclass prediction, you need to specify an average other than 'binary'


variance : Model variance by numpy.var()

outer_scores : A list of the outer scores, from the outer cross-validation

best_inner_score_list : A list of best inner scores for each outer loop

best_params : All best params from each inner loop cumulated in a dict

best_inner_params_list : Best inner params for each outer loop as an array of dictionaries

How to use the output?

We suggest looking at the best hyperparameters together with the score for each outer loop. Look at how stable the model appears to be in a nested cross-validation setting. If the outer score changes a lot, then it might indicate instability in your model. In that case, start over with making a new model.

After Nested Cross-Validation?

If the results from nested cross-validation are stable: Run a normal cross-validation with the same procedure as in nested cross-validation, i.e. if you used feature selection in nested cross-validation, you should also do that in normal cross-validation. Use the best parameters as input to your normal cross-validation.


  • XGBoost implements a early_stopping_rounds, which cannot be used in this implementation. Other similar parameters might not work in combination with this implementation. The function will have to be adopted to use special parameters like that.

What did we learn?

  • Using Scikit-Learn will lead to a faster implementation, since the Scikit-Learn community has implemented many functions that does much of the work.
  • We have learned and applied this package in our main project about House Price Prediction.

Why use Nested Cross-Validation?

Controlling the bias-variance tradeoff is an essential and important task in machine learning, indicated by [Cawley and Talbot, 2010]. Many articles indicate that this is possible by the use of nested cross-validation, one of them by Varma and Simon, 2006. Other interesting literature for nested cross-validation are [Varoquaox et al., 2017] and [Krstajic et al., 2014].


Nested cross-validation for unbiased predictions. Can be used with Scikit-Learn, XGBoost, Keras and LightGBM, or any other estimator that implements the scikit-learn interface.




No packages published