Study on Emergent Functional Evolution and Heuristic Knowledge Acquisition by Neural Structural Development

突现功能进化和神经结构发展启发式知识获取的研究

• 批准号: 17500083
• 负责人: HOMMA Noriyasu
• 金额: $ 1.28万
• 依托单位: Tohoku University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (C)
• 财政年份: 2005
• 资助国家: 日本
• 起止时间: 2005 至 2006
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-17500083/
• 关键词: Neural networks; Functional formation; Heuristic methods

In this research project, we analyze neural spike dynamics of a double feedback neural unit (DFNU) to clarify physiological microscopic mechanism of neural spike communication and functional formation. An essential emphasis of the analysis is on use of the DFNU's simple formulations that can provide quantitative analytic results. Comparing dynamics of Hodgkin-Huxley model to that of the DFNU, it is shown that dynamics of the DFNU is also physiologically plausible under a condition. The results suggest that high-frequency firings are relatively appropriate for a neural informational carrier due to the reliability and robustness to noisy inputs. To realize such reliable spike communication, we improved the DFNU's performance by using extra noisy inputs with appropriate amplitudes. Simulation studies show that there is optimal region of the amplitude that makes the DFNU possess the noise-enhanced reliable communication ability as similar to stochastic resonance phenomena.In addition to the microscopic analysis, we further conduct experimental analysis to investigate a key mechanism of emergent functional evolution from a viewpoint of macroscopic brain sciences. A phased reinforcement learning algorithm for controlling nonholonomic systems is proposed for this purpose. The key element of the proposed algorithm is a shaping function defined on a novel position-direction space. The shaping function is started to be constructed once the goal is reached and constrains the exploration strategy. The proposed method is applied to the positioning tasks of a 2-link planer underactuated manipulator. This manipulator has one passive joint and is difficult to control. As the result, the efficiency of the proposed shaping function was confirmed in learning speed and best policy.

本研究通过分析双反馈神经单元（DFNU）的神经尖峰动力学，阐明神经尖峰通讯和功能形成的生理微观机制。分析的一个基本重点是使用DFNU的简单公式，可以提供定量分析结果。将霍奇金-赫胥黎模型的动力学与DFNU模型的动力学进行比较，表明在一定条件下，DFNU的动力学在生理上也是合理的。结果表明，由于高频激发对噪声输入的可靠性和鲁棒性，它相对适合作为神经信息载体。为了实现这种可靠的尖峰通信，我们通过使用适当幅度的额外噪声输入来提高DFNU的性能。仿真研究表明，存在一个最优振幅区域，使得DFNU具有类似于随机共振现象的噪声增强可靠通信能力。在微观分析的基础上，我们进一步从宏观脑科学的角度进行实验分析，探讨突发性功能进化的关键机制。为此，提出了一种控制非完整系统的阶段性强化学习算法。该算法的关键要素是在新的位置-方向空间上定义的整形函数。一旦目标达到，就开始构造成形函数，并对勘探策略进行约束。将该方法应用于二连杆刨床欠驱动机械臂的定位任务。该机械手只有一个被动关节，控制困难。结果表明，所提出的成形函数在学习速度和最优策略方面是有效的。

项目成果

An analysis of the learning process on manual control of nonholonomic systems

非完整系统手动控制学习过程分析

• 发表时间: 2006
• 期刊: Japanese Journal of Ergonomics, (in Japanese) Vol.42, No.5
• 作者: Takahiro Tanaka;Yoshiaki Yasumura;Daisuke Katagami;Katsumi Nitta;Takakuni Goto;Norivasu Homma;Takakuni Goto;Noriyasu Homma;Takakuni Goto;後藤 太邦;Masao Sakai;後藤 太邦;Noriyasu Homma;Kenji Baheux;A-M.Zou;Masao Sakai;Takakuni Goto;Masao Sakai;Takakuni Goto
• 通讯作者: Takakuni Goto

非ホロノミック系における手動制御の学習過程に関する一考察

非完整系统手动控制学习过程研究

• 发表时间: 2006
• 期刊: 人間工学 42・5
• 作者: Takahiro Tanaka;Yoshiaki Yasumura;Daisuke Katagami;Katsumi Nitta;Takakuni Goto;Norivasu Homma;Takakuni Goto;Noriyasu Homma;Takakuni Goto;後藤 太邦;Masao Sakai;後藤 太邦;Noriyasu Homma;Kenji Baheux;A-M.Zou;Masao Sakai;Takakuni Goto;Masao Sakai;Takakuni Goto;Kenji Baheux;A-M Zou;M.M.Gupta;後藤 太邦
• 通讯作者: 後藤 太邦

Neural Spike Communication under Noisy Environments

• DOI: 10.1109/cibcb.2005.1594913
• 发表时间: 2005
• 期刊: 2005 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology
• 作者: N. Homma;K. Fuchigami;M. Gupta

Diagnosis and rehabilitation of hemispatial neglect patients with virtual reality technology

虚拟现实技术对偏侧空间忽视患者的诊断与康复

• 发表时间: 2005
• 期刊: Technology and Health Care 13・4
• 作者: Takahiro Tanaka;Yoshiaki Yasumura;Daisuke Katagami;Katsumi Nitta;Takakuni Goto;Norivasu Homma;Takakuni Goto;Noriyasu Homma;Takakuni Goto;後藤 太邦;Masao Sakai;後藤 太邦;Noriyasu Homma;Kenji Baheux;A-M.Zou;Masao Sakai;Takakuni Goto;Masao Sakai;Takakuni Goto;Kenji Baheux;A-M Zou;M.M.Gupta;後藤 太邦;Noriyasu Homma;後藤 太邦;Masao Sakai;Noriyasu Homma;Kenji Baheux
• 通讯作者: Kenji Baheux

Fuzzy Neural Networks

• DOI: 10.1016/0025-5564(75)90125-x
• 发表时间: 1975-02
• 期刊: Bellman Prize in Mathematical Biosciences
• 作者: Samuel C. Lee;Edward T. Lee