Construct mathematical models related to brain information processing and develop data analysis methods based on these models.

# Fumiyasu Komaki, Ph.D.

Unit Leader, Mathematical Informatics Collaboration Unit

fumiyasu.komaki [at] riken.jp

## Research Overview

Acquiring knowledge based on data in neuroscience is becoming increasingly important. By analyzing data using appropriate mathematical and statistical models, we extract information that is difficult to obtain by directly applying existing statistical methods. Techniques such as analyzing neural spike and brain wave data using point process models, decomposition of periodic components, and phase estimation in time series data using state space models, statistical learning with graph embedding, and information geometric methods are employed. In collaboration with researchers within and outside CBS, we advance research in mathematical informatics in neuroscience.

### Main Research Fields

- Complex Systems

### Related Research Fields

- Engineering
- Informatics
- Interdisciplinary Science & Engineering
- Mathematical & Physical Sciences
- Mathematical informatics
- Mathematical modelling
- Statistics

### Keywords

- Statistical modelling
- Information geometry
- Time series analysis
- Graphical model
- Statistical learning

## Selected Publications

- Komaki, F.

"Bayesian prediction and estimation based on a shrinkage prior for a Poisson regression model"

Japanese Journal of Statistics and Data Science (2024). - Kurata, S., Kuroda, R., and Komaki, F.

"Statistical modeling for temporal dominance of sensations data incorporating individual characteristics of panelists: an application to data of milk chocolate"

Journal of Food Science and Technology, 59, pp. 2420–2428 (2022). - Matsuda, T., Homae, F., Watanabe, H., Taga, G., and Komaki, F.

"Oscillator decomposition of infant fNIRS data"

PLoS Computational Biology, 18, e1009985 (2022). - Komaki, F.

"Shrinkage priors for nonparametric Bayesian prediction of nonhomogeneous Poisson processes"

IEEE Transactions on Information Theory, 2021, pp. 5305–5317, 2021 - Oda, H., and Komaki, F.

Shrinkage priors on complex-valued circular-symmetric autoregressive processes

IEEE Transactions on Information Theory, 67, pp. 5318–5333 (2021). - Yano, K., Kaneko, R., and Komaki, F.

Minimax predictive density for sparse count data

Bernoulli, 27, pp. 1212–1238 (2021). - Shibue, R., and Komaki, F.

Deconvolution of calcium imaging data using marked point processes

PLoS Computational Biology, 16, e1007650 (2020). - Shibue, R., and Komaki, F.

Firing rate estimation using infinite mixture models and its application to neural decoding

Journal of Neurophysiology, 118, pp. 2902–2913 (2017). - Matsuda, T., Kitajo, K., Yamaguchi, Y., and Komaki, F.

A point process modeling approach for investigating the effect of online brain activity on perceptual switching

NeuroImage, 152, pp. 50–59 (2017). - Matsuda, T., and Komaki, F.

Time series decomposition into oscillation components and phase estimation

Neural Computation, 29, pp. 332–367 (2017)

## Lab Members

### Principal investigator

- Fumiyasu Komaki
- Unit Leader

### Core members

- Xiyang Sun
- Postdoctoral Researcher
- Yuri Inoue
- Administrative Part-time Worker I