(A) simulation study on memory formation and maintenance using spiking neural network model스파이크 신경망 모델을 이용한 기억 형성 및 유지에 대한 시뮬레이션 연구

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 390
  • Download : 0
DC FieldValueLanguage
dc.contributor.advisorPaik, Se-Bum-
dc.contributor.advisor백세범-
dc.contributor.authorPark, Youngjin-
dc.date.accessioned2019-09-03T02:40:51Z-
dc.date.available2019-09-03T02:40:51Z-
dc.date.issued2018-
dc.identifier.urihttp://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=733819&flag=dissertationen_US
dc.identifier.urihttp://hdl.handle.net/10203/266169-
dc.description학위논문(석사) - 한국과학기술원 : 바이오및뇌공학과, 2018.2,[iv, 53 p. :]-
dc.description.abstractLearning new information and storing memories are two of the most important functions of the brain. The purpose of the thesis is to develop a computational model that accounts for the mechanism of memory formation and maintenance. To approach this question, we designed three simulations covering the circuit, molecular, and system levels, respectively. First, we discuss the roles of diverse spike-timing-dependent plasticity (STDP) profiles in flexible and stable memory formation. Through computer simulation, we found that an asymmetric learning rate generates flexible memory that is volatile and easily overwritten by newly appended information. Moreover, a symmetric learning rate generates stable memory that can coexist with newly appended information. Our results demonstrate that various attributes of memory functions may originate from differences in synaptic stability. Second, we suggested the roles of intrinsic excitability in memory allocation. Using the model network, we successfully generated experimental results that showed that a neuron excited just before learning has a higher probability of being allocated in a memory ensemble than control group neurons. This result suggests that temporal excitability fluctuation not only can link two memories encoded close in time, but also boost the capacity of the network. Finally, we present a novel systematic model in which the brain regions responsible for learning and coding are separated. We found that our model can account for the observed dynamics of an ensemble in the basolateral amygdala during fear conditioning. The simulated ensemble dynamics well described the observed data such as bi-directional plasticity, which never can be achieved by a single-layered network simulation. Overall, our thesis presents an integrated approach to memory formation and maintenance through computer simulation. We hope this study can provide a unified description of how and why our memories are formed and can be maintained.-
dc.languageeng-
dc.publisher한국과학기술원-
dc.subjectLearning and memory▼amemory model▼asynaptic plasticity▼aspike-timing-dependent plasticity▼aneural ensemble formation▼aneural excitability▼anon-Hebbian ensemble dynamics-
dc.subject학습과 기억▼a기억 모델▼a시냅스 가소성▼a스파이크 신호의 타이밍에 따른 가소성▼a뉴런 앙상블 형성▼a신경 흥분성▼a비헤비안 동역학-
dc.title(A) simulation study on memory formation and maintenance using spiking neural network model-
dc.title.alternative스파이크 신경망 모델을 이용한 기억 형성 및 유지에 대한 시뮬레이션 연구-
dc.typeThesis(Master)-
dc.identifier.CNRN325007-
dc.description.department한국과학기술원 :바이오및뇌공학과,-
dc.contributor.alternativeauthor박영진-
Appears in Collection
BiS-Theses_Master(석사논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0