Targeting strong predictors can significantly boost random forest regression accuracy
In short: Researchers found that pre-selecting the most relevant variables before running a random forest regression can greatly improve predictive performance. By focusing on strong predictors, this method helps the model avoid being weakened by irrelevant data in complex datasets, particularly when forecasting over long horizons.
The study used simulations with finite representative samples to demonstrate that targeting predictors helps control where the model places splits, making individual trees stronger. However, the researchers noted that targeting introduces a bias-variance trade-off because it increases the correlation between trees in the forest.
The authors found that selecting the best 5% to 30% of predictors balanced this trade-off, leading to predictive accuracy improvements of up to 21% in both economic recessions and expansions.
Why it matters: Focusing on the most relevant variables can dramatically improve the performance of random forest models.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Source: Research: Borup et al. (2022). Targeting predictors in random forest regression. International Journal of Forecasting. · cc-by-nc-nd
To cite this card, link to https://intellectplus.co.uk/app/science-research-targeting-strong-predictors-can-significantly-boost-random-f.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in Science · Latest · A–Z · All topics