1.
快速入门
1.
创建Notebook
2.
使用AWS Marketplace上的模型
3.
SageMaker Studio
4.
SageMaker Canvas
5.
SageMaker Studio架构
6.
特征工程
7.
训练和部署
8.
SageMaker 推理费用
9.
SageMaker Quota
10.
删除Domain CLI
2.
Data Wrangler - I
1.
数据准备
2.
Data Wrangler架构
3.
创建Data Flow
4.
Exploratory Data Analysis
5.
直方图和散点图
6.
Bias Detection - 偏差检测
7.
自定义可视化
8.
特征相关性
9.
Multicollinearity
3.
Data Wrangler - II
1.
检测重复行
2.
Quick Model
3.
特征工程 - 自定义转换
4.
特征工程 - 删除冗余列
5.
特征工程 - 处理缺失值和异常值
6.
特征工程 - 处理分类特征
7.
特征工程 - Normalize Numeric Features
8.
特征工程 - Balance Classes
9.
Quick Model
10.
导出 Data Wrangler Flow
11.
Programatic Export
5.
RTB Workshop
1.
环境搭建
2.
生成基本配置
3.
下载原始数据
4.
SageMaker Studio连接到 EMR 集群
5.
数据处理和特征工程
6.
探索数据
7.
模型训练
8.
推理模块
9.
附: 在SparkMagic上安装库
6.
Feature Store
1.
创
2.
AWS
3.
AWS
4.
4
5.
55
6.
AWS
7.
77
103.
Self-Paced Lab
104.
Module 1: Introduction to SageMaker Feature Store
105.
Prepare datasets
106.
Create feature group and ingest data
107.
Module 2: Working with offline store
108.
Working with offline store
109.
Search and Discovery using Feature-Level Metadata
110.
Speed ML Development with Apache Iceberg offline store compaction
111.
Module 3: Feature transformation and training
112.
Batch Scoring using a pre-trained XGBoost model
113.
Batch Scoring using a pre-trained XGBoost model (Parquet)
114.
Update Feature Group (Optional notebook)
115.
Feature transformation and training
116.
Module 4: Working with online store
117.
Working with online store
118.
Inference Patterns - endpoint based feature lookup
119.
Inference Patterns - Inference pipeline based feature lookup
120.
Module 5: Scalable Batch Ingestion
121.
Partition datasets
122.
Scikitlearn processor
123.
Apache Spark processor
124.
Feature Ingestion using Data Wrangler
125.
Module 6: Automated pipeline
126.
Module 7: Feature monitoring
127.
Module 8: ML Lineage Tracking
128.
Module 9: Security
129.
Granular access control in AWS Lake Formation (for Offline Feature Store)
130.
Access control to online and offline feature store using IAM policies
131.
Cleanup
清除历史
SageMaker介绍
>
Data Wrangler - III
>
标题：清理 权重：9000
标题：清理 权重：9000
删除 S3 中的工件。
删除 SageMaker Studio 中的数据流文件。
停止活动的 Data Wrangler 实例。
删除 SageMaker 用户配置文件和域（可选）。