Pawelzik, KlausAlbers, ChristianChristianAlbers2020-03-092020-03-092015-07-15https://media.suub.uni-bremen.de/handle/elib/896A central hypothesis of neuroscience is that the change of the strength of synaptic connections between neurons is the basis for learning in the animal brain. However, the rules underlying the activity dependent change as well as their functional consequences are not well understood. This thesis develops and investigates several different quantitative models of synaptic plasticity. In the first part, the Contribution Dynamics model of Spike Timing Dependent Plasticity (STDP) is presented. It is shown to provide a better fit to experimental data than previous models. Additionally, investigation of the response properties of the model synapse to oscillatory neuronal activity shows that synapses are sensitive to theta oscillations (4-10 Hz), which are known to boost learning in behavioral experiments. In the second part, a novel Membrane Potential Dependent Plasticity (MPDP) rule is developed, which can be used to train neurons to fire precisely timed output activity. Previously, this could only be achieved with artificial supervised learning rules, whereas MPDP is a local activity dependent mechanism that is supported by experimental results.eninfo:eu-repo/semantics/openAccessNeuronal networkssynaptic plasticitySTDPtheta oscillationslearningmemoryassociative learning530Dissertationurn:nbn:de:gbv:46-00104657-17