arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1706.00280cs.NE
收藏

回波状态网络:数字硬件的有效储备池计算

Integer Echo State Networks: Efficient Reservoir Computing for Digital Hardware

Denis Kleyko, E. Paxon Frady, Mansour Kheffache, Evgeny Osipov

摘要:

我们提出了一个近似的回声状态网络(ESN),可以有效地实现数字硬件的基础上的数学多维计算。所提出的整数回声状态网络(intESN)的库是仅包含n位整数的向量(其中n<8通常足以获得令人满意的性能)。递归矩阵乘法被替换为一个有效的循环移位操作。所提出的intESN的方法进行了验证,在储备池计算的典型任务:记忆的输入序列,分类时间序列,学习动态过程。这样的体系结构在内存占用和计算效率方面有了显著的改进,同时性能损失最小。现场可编程门阵列上的实验证实,所提出的intESN方法是更有效的能源比传统的ESN。

英文摘要:

We propose an approximation of Echo State Networks (ESN) that can be efficiently implemented on digital hardware based on the mathematics of hyperdimensional computing. The reservoir of the proposed integer Echo State Network (intESN) is a vector containing only n-bits integers (where n<8 is normally sufficient for a satisfactory performance). The recurrent matrix multiplication is replaced with an efficient cyclic shift operation. The proposed intESN approach is verified with typical tasks in reservoir computing: memorizing of a sequence of inputs; classifying time-series; learning dynamic processes. Such architecture results in dramatic improvements in memory footprint and computational efficiency, with minimal performance loss. The experiments on a field-programmable gate array confirm that the proposed intESN approach is much more energy efficient than the conventional ESN.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑