摘要
對於研究大腦的功能來說瞭解神經網路的構造與動力機制是很重要的，神經元是大腦中基本的組成單位，雖然在神經元的物理性質與行為的研究上已經吸引了相當大的注意，但是在神經網路中神經元之間的交互作用仍然不清楚。 在這研究中我們結合了Hodgkin-Huxley 的神經元模型與一個簡單地形成突觸聯結的機制來建構一個二維的神經網路，並且藉由STDP 學習機制使用電腦模擬來研究神經網路的同步化現象與突觸可塑性的性質變化。 在我們的模擬中，我們將藉由不同的分佈模式與分佈密度來探討神經網路發展時的性質。我們也採用最短路徑與叢集係數來探討網路的構造特性與現象。另外，我們也對神經網路在學習機制下所展現出的同步化行為的特徵進行研究。最後，我們將對於神經網路在發展的各階段經由學習效應的作用下所造成的突觸可塑性變化與增強效應進行探討。
並列摘要
Understanding the structure and dynamics of neural networks is important to study functions of brain. Neurons are the fundamental units of brains. Although the study in physical properties and behaviors of a single neuron has attracted considerable attention, the interaction among neurons in a neural network is so far unclear. In this study, we combine the Hodgkin-Huxley neural model and a simple mechanism of synaptic connection to build a two-dimension neural network, by which we could study the synchronization and synaptic plasticity of a neural network with STDP learning rule using computer simulations. In our simulations, we discuss the properties of a developing neural network with four designed patterns and different density of neural distribution. We also discuss the structural properties of networks by their characteristic path length and clustering coefficient. Furthermore, we investigate the synchronous behavior of a neural network after learning at various developing stages . Finally, we discuss the synaptic plasticity and enhancement effects for a random network after learning at various developing stages.
參考文獻
[23] Michael Haber and Keith K. Murai , Neuron Glia Biology , 59-66 (2006).
[2] Christof Koch , Biophysics of Computation – Information Processing in Single Neuron” New York : Oxford University Press, 142-171(1999)
[5] G.Q. Bi , H.X. Wang , Physio. Behav. 77 , 551-555 (2002) .
[6] G.Q. Bi , Biol. Cybern. 87 , 319-332 (2002) .
[7] G.Q. Bi , M.M Poo , Annu. Rev. Neurosci. 24, 139-166 (2001).
延伸閱讀
- 蔡政宏（2022）。視覺化模擬輔助人工智慧教學之研究-以類神經網路為例〔碩士論文，國立臺灣師範大學〕。華藝線上圖書館。https://doi.org/10.6345/NTNU202201807
- 蘇信華（2001）。類神經網路於模態識別之應用〔碩士論文，中原大學〕。華藝線上圖書館。https://doi.org/10.6840/cycu200100442
- 陳文峯（2017）。神經型態運算發展現況與展望。電工通訊季刊，()，49-57。https://doi.org/10.6328/CIEE.2017.3.06
- 李穎文（1990）。神經網路系統及其對立體影像計算之應用。光學工程，(35)，21-28。https://doi.org/10.30011/OE.199003.0003
- Carabez, E., Sugi, M., Nambu, I., & Wada, Y. (2017). Convolutional Neural Networks with 3D Input for P300 Identification in Auditory Brain-Computer Interfaces. Computational Intelligence and Neuroscience, 2017(), 1-9. https://doi.org/10.1155/2017/8163949