"Task-relevant autoencoding" enhances machine learning for human neuroscience

TL;DR

The paper introduces TRACE, a task-relevant autoencoder that improves extraction of behaviorally relevant neural representations from limited neuroimaging data, outperforming standard models in accuracy and clarity.

Contribution

We developed TRACE, a novel autoencoder that emphasizes task-relevant features, demonstrating superior performance over existing models in human neuroscience data analysis.

Findings

TRACE achieved up to 12% higher classification accuracy.

TRACE improved the discovery of task-relevant representations by up to 56%.

The model outperformed standard autoencoders and PCA on small datasets.

Abstract

In human neuroscience, machine learning can help reveal lower-dimensional neural representations relevant to subjects' behavior. However, state-of-the-art models typically require large datasets to train, so are prone to overfitting on human neuroimaging data that often possess few samples but many input dimensions. Here, we capitalized on the fact that the features we seek in human neuroscience are precisely those relevant to subjects' behavior. We thus developed a Task-Relevant Autoencoder via Classifier Enhancement (TRACE), and tested its ability to extract behaviorally-relevant, separable representations compared to a standard autoencoder, a variational autoencoder, and principal component analysis for two severely truncated machine learning datasets. We then evaluated all models on fMRI data from 59 subjects who observed animals and objects. TRACE outperformed all models nearly unilaterally, showing up to 12% increased classification accuracy and up to 56% improvement in discovering "cleaner", task-relevant representations. These results showcase TRACE's potential for a wide variety of data related to human behavior.