# Stabilizing Subject Transfer in EEG Classification with Divergence Estimation

arXiv: https://arxiv.org/abs/2310.08762 · HTML: https://mlsanity.com/paper/2310.08762/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 40 references that are in the map.
- First posted on arXiv 12 Oct 2023.

## Builds on (most cited first, 20 of 20)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Spectral Normalization for Generative Adversarial Networks](https://mlsanity.com/paper/1802.05957/) (2018)
- [On Variational Bounds of Mutual Information](https://mlsanity.com/paper/1905.06922/) (2019)
- [f-GAN: Training Generative Neural Samplers using Variational Divergence Minimization](https://mlsanity.com/paper/1606.00709/) (2016)
- [Learning Generative Models with Sinkhorn Divergences](https://mlsanity.com/paper/1706.00292/) (2017)
- [Telescoping Density-Ratio Estimation](https://mlsanity.com/paper/2006.12204/) (2020)
- [Wasserstein Dependency Measure for Representation Learning](https://mlsanity.com/paper/1903.11780/) (2019)
- [Transfer Learning for EEG-Based Brain-Computer Interfaces: A Review of Progress Made Since 2016](https://mlsanity.com/paper/2004.06286/) (2020)
- [Transfer Learning in Brain-Computer Interfaces](https://mlsanity.com/paper/1512.00296/) (2015)
- [Understanding and Improving the Role of Projection Head in Self-Supervised Learning](https://mlsanity.com/paper/2212.11491/) (2022)
- [Neural Estimation of Statistical Divergences](https://mlsanity.com/paper/2110.03652/) (2021)
- [Practical and Consistent Estimation of f-Divergences](https://mlsanity.com/paper/1905.11112/) (2019)
- [Bayes-Ball: The Rational Pastime (for Determining Irrelevance and Requisite Information in Belief Networks and Influence Diagrams)](https://mlsanity.com/paper/1301.7412/) (2013)
- [Adversarial Symmetric Variational Autoencoder](https://mlsanity.com/paper/1711.04915/) (2017)
- [Transfer Learning in Brain-Computer Interfaces with Adversarial Variational Autoencoders](https://mlsanity.com/paper/1812.06857/) (2018)
- [Adversarial Deep Learning in EEG Biometrics](https://mlsanity.com/paper/1903.11673/) (2019)
- [Invariant Representations from Adversarially Censored Autoencoders](https://mlsanity.com/paper/1805.08097/) (2018)
- [Disentangled Adversarial Transfer Learning for Physiological Biosignals](https://mlsanity.com/paper/2004.08289/) (2020)
- [AutoTransfer: Subject Transfer Learning with Censored Representations on Biosignals Data](https://mlsanity.com/paper/2112.09796/) (2021)
- [Recursive Estimation of User Intent from Noninvasive Electroencephalography using Discriminative Models](https://mlsanity.com/paper/2211.02630/) (2022)

## Abstract (arXiv)

Classification models for electroencephalogram (EEG) data show a large decrease in performance when evaluated on unseen test sub jects. We reduce this performance decrease using new regularization techniques during model training. We propose several graphical models to describe an EEG classification task. From each model, we identify statistical relationships that should hold true in an idealized training scenario (with infinite data and a globally-optimal model) but that may not hold in practice. We design regularization penalties to enforce these relationships in two stages. First, we identify suitable proxy quantities (divergences such as Mutual Information and Wasserstein-1) that can be used to measure statistical independence and dependence relationships. Second, we provide algorithms to efficiently estimate these quantities during training using secondary neural network models. We conduct extensive computational experiments using a large benchmark EEG dataset, comparing our proposed techniques with a baseline method that uses an adversarial classifier. We find our proposed methods significantly increase balanced accuracy on test subjects and decrease overfitting. The proposed methods exhibit a larger benefit over a greater range of hyperparameters than the baseline method, with only a small computational cost at training time. These benefits are largest when used for a fixed training period, though there is still a significant benefit for a subset of hyperparameters when our techniques are used in conjunction with early stopping regularization.