Tomoyuki Hiroyasu

This study introduces a robust cooperative learning method for multiview regression that reduces outlier influence while preserving view information sharing, showing improved predictive performance in labor onset prediction.

This study compares deep generative models for generating brain structural connectivity data, finding that WGAN-GP provides the most stable performance while VAE and DDPM have application-specific strengths.

This paper introduces ELA/GAopt, a framework using genetic algorithms to automate ROI selection in brain dynamics analysis, showing improved reproducibility and condition-specific insights in neuroimaging data.

The paper introduces Task-guided GAN II to augment structural connectivity matrices for connectome-based prediction, showing improved accuracy with limited real data.

This study uses fNIRS to compare hemispheric lateralization between expert and non-expert dart players, finding higher lateralization in experts and younger individuals.

The paper compares classification accuracy and feature selection between sparse and non-sparse modeling of metabolomics data using serum lipidomic data from breast cancer patients.

The paper proposes new treatment effect estimation methods using L0 norm-based lasso and ridge regression to identify subgroups of patients who may respond to a new treatment.

This study examines brain synchronization during economic exchanges between acquainted pairs, finding no increase in IBS compared to strangers but identifying pair-specific synchronization in specific brain regions during the task.

This study develops a model to predict distracted driving using fNIRS-measured brain activity, finding that specific neural network connections are key predictors of distraction levels.

This study evaluates sparse k-means for metabolomics data, showing it improves clustering accuracy with fewer metabolites in breast cancer subtypes.

This study uses Tucker3 clustering on graph theoretical metrics to show that breath-counting meditation increases clustering coefficients in eight brain regions, indicating network density changes.