A Metamodel Structure For Regression Analysis: Application To Prediction Of Autism Spectrum Disorder Severity

TL;DR

This paper introduces a novel metamodel structure combining classification and regression models to enhance prediction accuracy and generalizability in small, noisy datasets, demonstrated on autism severity prediction from fMRI data.

Contribution

A new metamodel architecture that integrates multiple base classifiers with a regression model, improving performance in small, noisy data scenarios.

Findings

Metamodel outperforms traditional regression models in accuracy and stability.

Metamodel is more flexible and generalizable.

Effective in predicting autism severity from fMRI data.

Abstract

Traditional regression models do not generalize well when learning from small and noisy datasets. Here we propose a novel metamodel structure to improve the regression result. The metamodel is composed of multiple classification base models and a regression model built upon the base models. We test this structure on the prediction of autism spectrum disorder (ASD) severity as measured by the ADOS communication (ADOS COMM) score from resting-state fMRI data, using a variety of base models. The metamodel outperforms traditional regression models as measured by the Pearson correlation coefficient between true and predicted scores and stability. In addition, we found that the metamodel is more flexible and more generalizable.