Aggregating density estimators: an empirical study

TL;DR

This paper introduces new density estimation algorithms based on bootstrap aggregation, compares them empirically with existing methods, and demonstrates their effectiveness and computational efficiency in density estimation tasks.

Contribution

It presents novel ensemble-based density estimators and provides an empirical comparison with established methods like stacking and boosting.

Findings

Ensemble methods are effective for density estimation.

Some proposed algorithms are simpler and more computationally efficient.

Ensemble methods can perform comparably to traditional density estimation techniques.

Abstract

We present some new density estimation algorithms obtained by bootstrap aggregation like Bagging. Our algorithms are analyzed and empirically compared to other methods found in the statistical literature, like stacking and boosting for density estimation. We show by extensive simulations that ensemble learning are effective for density estimation like for classification. Although our algorithms do not always outperform other methods, some of them are as simple as bagging, more intuitive and has computational lower cost.