Meta-Ensemble Learning with Diverse Data Splits for Improved Respiratory Sound Classification

TL;DR

This paper introduces a meta-ensemble learning approach that trains base models on diverse data splits to improve respiratory sound classification accuracy and generalization, achieving state-of-the-art results.

Contribution

It proposes a novel meta-ensemble method that enhances prediction diversity by using different data splits, leading to better performance on respiratory sound datasets.

Findings

Achieved a new state-of-the-art score of 66.49% on the ICBHI benchmark.

Improved generalization demonstrated on out-of-distribution datasets.

Training on diverse data splits enhances ensemble robustness and accuracy.

Abstract

Training reliable respiratory sound classification models remains challenging due to the limited size and subject diversity of datasets. Ensemble methods can improve robustness, but when base models are trained on identical data, models tend to overfit and produce highly correlated predictions, thereby reducing the effectiveness of ensembling. In this work, we investigate a meta-ensemble learning methodology that enhances prediction diversity by training base models on diverse data splits and combining their outputs through a trained meta-model. Specifically, we train base models on the ICBHI dataset using two data split settings: fixed 80-20% split and five-fold cross-validation split, under two data granularity settings: patient- and sample-level. The resulting diversity in base model predictions enables the meta-model to better generalize. Our approach achieves new state-of-the-art performance on the ICBHI benchmark, reaching a Score of 66.49% and showing improved generalization on two out-of-distribution datasets, indicating its potential applicability to real-world clinical data.