Machine learning (ML) is adopted into a wide range of applications. The popularity and adoption came because of performance and capability to deploy in real-world systems and the ease of use. Given a wide range of works proposed on ML, ranging from the ML learning methodology to the design of hardware accelerators, different works focus on various aspects, depending on the application requirements. For a fair comparison and benefit of the society, benchmarking of the ML applications and performance is nontrivial. Such a benchmarking will also enable the community to analyze and fairly compare the solutions.

中文翻译：

---

客座编辑介绍：机器学习系统和应用基准测试特刊


机器学习（ML）被广泛应用。它的流行和采用是因为在现实系统中部署的性能和能力以及易用性。鉴于机器学习方面提出的工作范围广泛，从机器学习学习方法到硬件加速器的设计，不同的工作根据应用需求侧重于不同的方面。为了公平比较并造福社会，对机器学习应用程序和性能进行基准测试并非易事。这样的基准测试还将使社区能够分析和公平地比较解决方案。