Kleyko Denis, Rahimi Abbas, Rachkovskij Dmitri A, Osipov Evgeny, Rabaey Jan M
IEEE Trans Neural Netw Learn Syst. 2018 Dec;29(12):5880-5898. doi: 10.1109/TNNLS.2018.2814400. Epub 2018 Apr 5.
Hyperdimensional (HD) computing is a promising paradigm for future intelligent electronic appliances operating at low power. This paper discusses tradeoffs of selecting parameters of binary HD representations when applied to pattern recognition tasks. Particular design choices include density of representations and strategies for mapping data from the original representation. It is demonstrated that for the considered pattern recognition tasks (using synthetic and real-world data) both sparse and dense representations behave nearly identically. This paper also discusses implementation peculiarities which may favor one type of representations over the other. Finally, the capacity of representations of various densities is discussed.
超维（HD）计算是未来低功耗运行的智能电子设备的一种很有前景的范式。本文讨论了将二进制HD表示应用于模式识别任务时选择参数的权衡。具体的设计选择包括表示的密度以及从原始表示映射数据的策略。结果表明，对于所考虑的模式识别任务（使用合成数据和真实世界数据），稀疏表示和密集表示的表现几乎相同。本文还讨论了可能使一种表示类型优于另一种表示类型的实现特点。最后，讨论了各种密度表示的容量。
