Why Random Forest is better than Decision tree
In this article I’m pointing some features of random forest that make it better model
1 min readMay 14, 2021
- Random forest is an ensemble model which combine multiple models(decision trees) in parallel and do the final prediction. Which reduce the over-fitting issue.
- Random in random forest means it will fit multiple trees in the random samples of the data(bootstrap samples).This bootstrap samples (sampling with replacement) reduce the variance in the data.
- Since random forest model using random samples each tree will fit on different set of data samples(hence dominant of a strong predictor variable will not be there).So the trees are decorrelated to each other.
- Random forest is faster in training as compared to decision tree since it divides the total data into subsamples
- Random forest is using bootstrap samples, so only 60% data will be used for training rest 30% is called oob (out of bag) samples which can be used for testing
I hope this help you !!! Happy learning