首页> 外文会议>International Conference on Artificial Neural Networks >Episodic Memory and Cognitive Map in a Rate Model Network of the Rat Hippocampus
【24h】

Episodic Memory and Cognitive Map in a Rate Model Network of the Rat Hippocampus

机译:大鼠海马速率模型网络中的焦虑记忆和认知地图

获取原文

摘要

The hippocampus of rodents seems to play a central role in navigation: place cells are prone to serve as elements of a cognitive map. The hippocampus of primates is apparently involved in episodic memory processes. Here we show how the hippocampi of different species (with their anatomy and physiology being considerably similar) can fulfill both purposes. A network was built from simple rate neurons, synaptic weights were modified by a batch version of the coincidence-based Bienenstock-Cooper-Munro learning rule. The network was trained with temporally contiguous sensory stimuli, and then tested in two different paradigms: (1) for navigational performance, place field profiles were evaluated to allow comparison with experimental data; (2) for memory performance, associative capability was assessed. The same neuron network was found to be able to show place cell properties and function as episodic memory, even when trained with non-topographical stimuli.
机译:啮齿动物的海马似乎在导航中发挥着核心作用:将小区倾向于作为认知地图的元素。灵长类动物的海马显然涉及焦虑内存过程。在这里,我们展示了不同物种的海马(其解剖学和生理学相当相似的病理学)是如何实现的。从简单的速率神经元建造了网络,突触权重通过批量版的基于巧合的Bienenstock-Cooper-Munro学习规则来修改。网络接受过时间连续的感觉刺激训练,然后在两个不同的范式中测试:(1)用于导航性能,评估PRIVITION字段配置文件以允许与实验数据进行比较; (2)对于内存性能,评估关联能力。也发现相同的神经元网络能够显示将细胞性质和功能作为情节内存器,即使在具有非地形刺激的训练中也是如此。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号