• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International conference on artificial intelligence >A novel biologically plausible supervised learning method for spiking neurons
【24h】

A novel biologically plausible supervised learning method for spiking neurons

机译：一种新的生物学上合理的监督神经元学习方法

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

A novel learning rule, Cross-Correlated Delay Shift (CCDS) learning algorithm, is proposed for processing spatiotemporal patterns in this study. CCDS is a supervised learning rule that is able to learn association of arbitrary spike trains in a supervised fashion. Single spiking neuron trained according to CCDS algorithm is capable of learning and precisely reproducing arbitrary target sequences of spikes. Unlike the ReSuMe learning rule, synapse delays and axonal delays in CCDS are variants which are modulated together with weights during learning. Besides biological plausibility, CCDS is also computationally efficient. In the presented experimental analysis, the proposed learning algorithm is evaluated by it properties including its robustness in dealing with noisy environment, and its adaptive learning performance to different spatio-temporal patterns. Simulation results have shown that the proposed CCDS learning method achieves learning accuracy and learning speed improvements comparable to ReSuMe.
机译：在这项研究中，提出了一种新颖的学习规则，即交叉相关延迟移位（CCDS）学习算法，用于处理时空模式。 CCDS是一种有监督的学习规则，它能够以有监督的方式学习任意峰值列车的关联。根据CCDS算法训练的单尖峰神经元能够学习并精确地复制任意的尖峰目标序列。与ReSuMe学习规则不同，CCDS中的突触延迟和轴突延迟是在学习过程中与权重一起调制的变体。除了生物学上的合理性外，CCDS的计算效率也很高。在提出的实验分析中，对所提出的学习算法进行了性能评估，包括其在嘈杂环境中的鲁棒性以及对不同时空模式的自适应学习性能。仿真结果表明，所提出的CCDS学习方法可以实现与ReSuMe相当的学习准确性和学习速度的提高。

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