Continuous Integration of Machine Learning Models with ease.ml/ci: Towards a Rigorous Yet Practical Treatment
Continuous Integration of Machine Learning Models with ease.ml/ci: Towards a Rigorous Yet Practical Treatment
复制标题
轻松持续集成机器学习模型.ml/ci：迈向严格而实用的治疗
DOI：
--
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
Ce Zhang
中科院分区：
文献类型：
--
作者：
Cédric Renggli;Bojan Karlas;Bolin Ding;Feng Liu;K. Schawinski;Wentao Wu;Ce Zhang
持续集成是现代软件工程实践中系统地管理系统开发生命周期的一个不可或缺的步骤。开发机器学习模型没有区别-它是一个具有生命周期的工程过程，包括设计，实现，调优，测试和部署。然而，大多数（如果不是全部的话）现有的持续集成引擎并不支持机器学习作为一等公民。
在本文中，我们提出了这个http URL，据我们所知，第一个机器学习的持续集成系统。构建此http URL的挑战是提供严格的保证，例如，具有0.999可靠性的单精度点误差容限，具有实际量的标记工作，例如，每次测试2K个标签。我们设计了一个域特定的语言，允许用户指定的集成条件与可靠性的约束，并开发简单的新颖的优化，可以降低标签的数量，所需的测试条件普遍使用的真实的生产系统的数量级高达两个。
Continuous integration is an indispensable step of modern software engineering practices to systematically manage the life cycles of system development. Developing a machine learning model is no difference - it is an engineering process with a life cycle, including design, implementation, tuning, testing, and deployment. However, most, if not all, existing continuous integration engines do not support machine learning as first-class citizens.
In this paper, we present this http URL, to our best knowledge, the first continuous integration system for machine learning. The challenge of building this http URL is to provide rigorous guarantees, e.g., single accuracy point error tolerance with 0.999 reliability, with a practical amount of labeling effort, e.g., 2K labels per test. We design a domain specific language that allows users to specify integration conditions with reliability constraints, and develop simple novel optimizations that can lower the number of labels required by up to two orders of magnitude for test conditions popularly used in real production systems.