|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/designing-ml-solutions-on-azure-preparing-for-dp-100-exam/
课程评论：没有评论
**Azure ML解决方案设计与DP-100认证准备课程概览（中文总结）** 本课程是一门全面的Azure机器学习实战指南，旨在帮助专业人士将机器学习模型从实验阶段成功部署到生产环境。课程涵盖了机器学习生命周期的各个阶段，并充分利用了Azure强大的工具套件。 **学习收获：** * **Azure ML架构设计：** 学习如何在Azure上设计有效的机器学习架构。 * **数据与计算：** 选择合适的数据集格式和计算目标。 * **实验优化：** 训练可扩展、高性能的机器学习实验。 * **DevOps集成：** 集成Git和CI/CD管道，实现高效协作。 * **数据管理：** 规模化准备和管理数据，利用Synapse Spark进行数据处理和转换。 * **模型开发：** 通过Azure ML Datastores访问和版本化数据集，构建和共享环境。 * **训练方法：** 掌握自动化（AutoML - 分类、回归、视觉、NLP）和自定义（Python、MLflow）模型训练方法。 * **超参数调优：** 优化模型性能。 * **MLOps实践：** 构建可复现的机器学习管道，实现模块化训练和数据传递。 * **部署策略：** 部署模型以进行实时和批量推理，配置在线和批量端点，确保安全合规。 * **高级AI与LLM：** 优化高级AI模型和大规模语言模型（LLMs），进行模型选择、微调，设计提示工程，并实现检索增强生成（RAG）。 * **负责任AI与运维：** 应用公平性、透明性和可解释性原则，利用MLflow进行实验跟踪和模型治理，实现自动化再训练和生产监控。 **目标受众：** 本课程特别适合希望扩展数据科学能力、准备DP-100认证考试、或希望提升组织AI能力的专业人士。课程提供了与业界标准相符的实践经验，帮助学员解决实际业务问题。
Build and Deploy Intelligent Machine Learning Solutions Using Microsoft AzureThis course is your complete guide to mastering data science workflows in the cloud. Designed for professionals who want to go beyond experimentation and take their machine learning models into production, it covers every stage of the ML lifecycle using Azure's powerful suite of tools.Whether you're looking to scale your data science capabilities, prepare for the DP-100 certification, or enhance your organization's AI capabilities, this course delivers hands-on experience with the platforms and practices used in real-world enterprise environments.You will gain hands-on expertise in:Designing effective ML architectures on AzureChoosing the right dataset formats and compute targetsStructuring experiments for scalability and performanceIntegrating Git and CI/CD pipelines for streamlined collaborationPreparing and managing data at scaleWrangling and transforming data using notebooks and Synapse SparkAccessing and versioning datasets via Azure ML datastoresBuilding and sharing environments across workspacesTraining models using both automated and custom approachesLeveraging AutoML for classification, regression, vision, and NLPDeveloping custom training scripts using Python and MLflowTuning hyperparameters for optimal model performanceBuilding and managing reproducible ML pipelinesCreating modular training componentsPassing and transforming data between pipeline stepsScheduling, monitoring, and debugging workflowsDeploying models for real-time and batch inferenceConfiguring online endpoints for scalable predictionsSetting up batch endpoints for large-scale processing jobsImplementing secure and compliant deployment workflowsOptimizing advanced AI models and LLMsSelecting and fine-tuning large language modelsDesigning prompt engineering strategies for accuracy and contextImplementing Retrieval Augmented Generation (RAG) systemsEnsuring responsible AI and operational excellenceApplying fairness, transparency, and explainability principlesUsing MLflow for experiment tracking and model governanceAutomating retraining and monitoring in productionIf you're ready to move beyond theory and start building machine learning systems that solve real business problems, this course is designed for you. It's perfect for learners who want structured guidance, practical tools, and hands-on labs that mirror what professionals do in industry every day.
