If I fit two RandomForestClassifiers 500 trees each and average their predicted probabilities on the test set, would it have better results than fitting a RandomForestClassifier with 1000 trees and use it to get test set probabilities?

As these algorithms are random based I would say that their performance should be roughly the same?

I am okay with some math to prove it, or any other way that might prove it.

  • $\begingroup$ Have you tried with a toy dataset? $\endgroup$
    – Multivac
    Oct 5, 2020 at 18:51

2 Answers 2


Short answer: they are equivalent.

Any results that suggest otherwise are due to random chance or due to modification of parameters other than the number of trees. A random forest is just a voted ensemble of decision trees. By default, each tree's vote is weighted equally and then these votes are averaged. Suppose sets X and Y are the same size. Then if you take the average of X, the average of Y, then average those two, that's the exact same thing as just combining X and Y and averaging them. This is the same as if you have two random forests with the same number of trees.

Note, however, that if they have a different number of trees, then if you build two forests and average them, the individual trees making up the smaller forest will have their votes weighted more highly than the trees in the larger forest.


With a toy dataset, I obtained slightly better results with the RF 1000 trees

import pandas as pd
import numpy as np
from sklearn.datasets import load_breast_cancer
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import roc_auc_score

from scipy.stats import ks_2samp

import matplotlib.pyplot as plt

n_models = 2
threshold = .5

X, y = load_breast_cancer(return_X_y= True)
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state = 42, test_size = .3)

rForest1000 = RandomForestClassifier(n_estimators= 1000, random_state= 42, oob_score= True).fit(X_train, y_train)
preds1000 = rForest1000.predict_proba(X_test)[:,1]

roc_score = roc_auc_score(y_true = y_test, y_score= preds1000)

print(f"Test set score for 1K trees is : {round(roc_score, 4)}")

ls = list()
for i in range(n_models):
    model = RandomForestClassifier(n_estimators= 500, random_state= i).fit(X_train, y_train)
preds = np.array(ls).mean(axis =0)

roc_score = roc_auc_score(y_true = y_test, y_score= preds)

print(f"Test set score for 500trees X 2 avg is : {round(roc_score, 4)}")

fig, ax = plt.subplots(1,2, figsize = (12,5))

ax[0].hist(preds1000[y_test == 0], color = "darkgreen", alpha = .5)
ax[0].hist(preds1000[y_test == 1], color = "darkred", alpha = .5)

ax[0].set_title(f"Predictions distribution with 1K Trees RF\n KS: {[round(ks_2samp(preds1000[y_test == 0], preds1000[y_test == 1])[0],3)]}")

ax[1].hist(preds[y_test == 0], color = "darkgreen", alpha = .5)
ax[1].hist(preds[y_test == 1], color = "darkred", alpha = .5)

ax[1].set_title(f"Predictions distribution with X2 500 Trees RF Trees RF\n KS: {[round(ks_2samp(preds[y_test == 0], preds[y_test == 1])[0],3)]}");
  • $\begingroup$ If you rerun this with different random seeds and analyze the resulting score distributions, I think you'll find there is no statistically significant difference. $\endgroup$ Oct 5, 2020 at 19:57
  • $\begingroup$ @BenReiniger I completely agree with you. Nonetheless, I have seen the second approach of averaging RF predictions on Kaggle competitions, in my opinion at the end is a matter of experimentation especially when a .001% improvement on performance is the difference between the 1st and the 11th place on a Kaggle contest, but in practical terms, there is no difference $\endgroup$
    – Multivac
    Oct 5, 2020 at 20:23
  • 1
    $\begingroup$ If the two RF's have different hyperparameters, then it's a different story (and that includes kaggle "ensemble-all-the-public-models-together" types). But if they have the same hyperparameters, then it's equivalent to the more-tree version, up to the random effects: you can't know in advance which one will come up on top for an unseen dataset. $\endgroup$ Oct 6, 2020 at 14:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.