2022 Fiscal Year Final Research Report
Theoretical research on spatiotemporal information processing in the brain
| Project/Area Number |
19K20365
|
|---|
| Research Category |
Grant-in-Aid for Early-Career Scientists
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
Basic Section 61040:Soft computing-related
|
|---|
| Research Institution | The University of Tokyo (2021-2022)
Institute of Physical and Chemical Research (2019-2020) |
|---|
Principal Investigator |
Terada Yu 東京大学, 大学院理学系研究科(理学部), 客員共同研究員 (40815338)
|
|---|
| Project Period (FY) |
2019-04-01 – 2023-03-31
|
| Keywords | リカレントニューラルネットワーク / 神経データ / 結合推定 / 非線形動力学 / カオス |
|---|
| Outline of Final Research Achievements |
In this research project, we develop statistical inference methods for reconstrcucting neuronal couplings from spiking data and study theoretical/computational models for neural dynamics and computations. Our methods were shown to exhibit high accuracny and efficiency in inferring synaptic couplings. We also bulit a linear response theory in coupled oscillators to study the properties of macroscopic and mesoscopic activity in the brain.We applied the theory to develop an inference method in coupled oscillators. Finally, we study the computational roles of deterministic activity in neural networks and demonstrated that the choatic neural activity may serve as a neural substrate for reprepsentations of probability distributions.
|
| Free Research Field |
理論神経科学
|
| Academic Significance and Societal Importance of the Research Achievements |
提案した結合推定手法は現状得られているデータだけでなく今後得られるよりサイズの大きいデータに対しても適用が期待される．神経回路網のモデルは神経科学の問題だけでなく，機械学習やAIといった関連分野への波及効果も期待できる．
|
