|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/using-sagemaker-pipelines-get-ml-models-approved-and-deploy/
课程评论：没有评论
**课程名称：使用 SageMaker Pipelines 实现 ML 模型审批与部署** **课程概述：** 本课程将教授您如何利用 Amazon SageMaker Pipelines 实现机器学习模型的审批与部署，让您在 MLOps（机器学习运维）领域更加得心应手。课程内容涵盖了 MLOps 的核心概念、SageMaker Studio 的实用功能，以及模型监控等关键环节。 **核心内容：** * **XGBoost 模型实践：** 课程包含四项实践练习，涉及三种类型的 XGBoost 模型：回归、二分类和多分类。您将学习如何构建、训练和评估这些模型。 * **模型审批流程：** 前两项练习将重点关注如何成功审批回归和二分类模型，使其能够进入生产环境。 * **模型部署与预测：** 后两项练习将指导您如何将模型部署到生产环境，并进行实际预测。 * **SageMaker Pipelines 五大步骤：** 深入讲解 SageMaker Pipelines 的五个关键步骤，帮助您构建高效、可重复的 ML 工作流。 * **深入机器学习知识：** 课程将进一步探讨机器学习的相关概念，提升您的理论和实践能力。 * **交叉验证精讲：** 详细讲解交叉验证技术，增强您对模型泛化能力的理解，从而提高模型审批的信心。 * **MLOps 概念与应用：** 强调 MLOps 的重要性，解释其如何解决模型漂移等行业痛点，并展示其在各行业的应用。 * **SageMaker Studio 功能探索：** 熟悉 SageMaker Studio 的各项功能，包括其在数据准备、模型训练、部署和监控中的作用。 * **模型监控：** 学习如何监控已部署模型的性能，及时发现并处理潜在问题。 **学习目标：** * 掌握使用 SageMaker Pipelines 构建和管理 ML 工作流的能力。 * 熟悉 XGBoost 在不同任务（回归、分类）中的应用。 * 了解并实践 ML 模型的审批和部署流程。 * 加深对 MLOps 概念和重要性的理解。 * 提升在机器学习和 Amazon SageMaker 方面的技能。 * 对模型监控有清晰的认识。 **预备知识：** * 具备基础的 MLOps 概念知识（非必须，课程中会讲解）。 * 对机器学习和 Amazon SageMaker 不要求是专家级别，课程会进行涵盖。 **推荐：** * 如果您已完成讲师的其他相关课程，本课程作为其续篇，将提供更深入的 MLOps 和 SageMaker Pipelines 内容。 * 如果您想学习如何在 AWS 上部署各种 SageMaker 模型，讲师的另一门课程是很好的补充，但该课程不包含 MLOps 和 SageMaker Pipelines 的深入内容。 **评估方式：** 课程包含四道测验题，旨在巩固您对课程内容的理解，难度适中，确保您能通过观看视频和完成练习来掌握知识。 **这是一门关于如何自信地将 ML 模型投入生产的实践课程，让我们一起愉快地学习吧！**
In this course there are 4 exercises of 3 different types of XGBoost models which are regression, binary classification, and multi class classification. The first two exercises you will get those two models approved for production. Then in the other two videos you will do the same on deploy and make predictions as well. In this video we will cover the 5 necessary pipeline steps and also get more in depth into more machine learning. Not to mention we will cover cross validation in depth and become more confident in getting models approved for production and a better understanding of MLOps. Also the workflow structure as well and learn features of Sagemaker Studio. Do not worry about having slight MLOps knowledge and not being an expert in Machine Learning or Amazon Sagemaker we will cover all of that including monitoring models as well. I will also have 4 quiz questions down below that will not be too easy or to difficult more of making sure that you watched the videos and did the exercises in the videos. But most importantly have fun learning. Don't forget that MLOps is very important in every Data Science project used in every industry because it addresses a common problem of model drift. If you have taken my other course I suggest you take this one as well because this is more of a sequel to the other one. Also if you are taking this course I suggest you take my other course to show you how to deploy various Sagemaker Models on AWS. But the other course does not include MLOps and Sagemaker Pipelines like this one does.
