大数据的发展已经引领了对能够学习包含数百万至数十亿参数的复杂模型的机器学习系统的新需求，以保证足够的能力来消化海量的数据集，提供强大的预测分析（如高维潜特征、中介表示和决策功能）。为了在这样的尺度上，在成百上千台的分布式机器集群中运行机器学习算法，关键往往是要投入显著的工程性的努力——有人可能会问，这样的工程是否还属于机器学习的研究领域？考虑到如此'大'的机器学习系统可以极大地从根植于机器学习的统计和算法的理解中受益——因此，机器学习的研究人员应该不会回避这样的系统设计——我们讨论了一系列从我们近来对工程尺度的机器学习解决方案的研究中提炼的原则和策略。这些原则和策略从机器学习的应用连续跨越到它的工程和理论研究，以及大型机器学习的系统和架构的发展，目标是了解如何使其有效、广泛地适用，并以收敛和缩放保证支持。它们关注的是机器学习研究传统上注意较少的四个关键问题：一个机器学习程序怎样能分布到一个集群中去？机器学习计算怎样能通过机器间的交流连接起来？这样的交流是如何被执行的？机器间应该交流的内容是什么？通过揭示机器学习程序所独有的，而非常见于传统计算机程序中的基础性的统计和算法上的特点，并通过剖析成功案例，以揭示我们如何利用这些原则来同时设计和开发高性能的分布式机器学习软件以及通用的机器学习框架，我们为机器学习的研究人员和从业者提供了进一步塑造并扩大机器学习与系统之间的领域的机会。

The rise of big data has led to new demands for machine learning (ML) systems to learn complex models, with millions to billions of parameters, that promise adequate capacity to digest massive datasets and offer powerful predictive analytics (such as high-dimensional latent features, intermediate representations, and decision functions) thereupon. In order to run ML algorithms at such scales, on a distributed cluster with tens to thousands of machines, it is often the case that significant engineering efforts are required—and one might fairly ask whether such engineering truly falls within the domain of ML research. Taking the view that 'big' ML systems can benefit greatly from ML-rooted statistical and algorithmic insights—and that ML researchers should therefore not shy away from such systems design—we discuss a series of principles and strategies distilled from our recent efforts on industrial-scale ML solutions. These principles and strategies span a continuum from application, to engineering, and to theoretical research and development of big ML systems and architectures, with the goal of understanding how to make them efficient, generally applicable, and supported with convergence and scaling guarantees. They concern four key questions that traditionally receive little attention in ML research: How can an ML program be distributed over a cluster? How can ML computation be bridged with inter-machine communication? How can such communication be performed? What should be communicated between machines? By exposing underlying statistical and algorithmic characteristics unique to ML programs but not typically seen in traditional computer programs, and by dissecting successful cases to reveal how we have harnessed these principles to design and develop both high-performance distributed ML software as well as general-purpose ML frameworks, we present opportunities for ML researchers and practitioners to further shape and enlarge the area that lies between ML and systems.