AWS Certified Machine Learning - Specialty (MLS-C01) Certification Guide - Second Edition: The ultimate guide to passing the MLS-C01 exam on your first attempt

Cover
FM
Copyright
Contributors
Table of Contents
Preface
Chapter 1: Machine Learning Fundamentals
	Making the Most Out of this Book – Your Certification and Beyond
	Comparing AI, ML, and DL
		Examining ML
		Examining DL
	Classifying supervised, unsupervised, and reinforcement learning
		Introducing supervised learning
	The CRISP-DM modeling life cycle
	Data splitting
		Overfitting and underfitting
		Applying cross-validation and measuring overfitting
		Bootstrapping methods
		The variance versus bias trade-off
		Shuffling your training set
	Modeling expectations
	Introducing ML frameworks
	ML in the cloud
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 2: AWS Services for Data Storage
	Technical requirements
	Storing Data on Amazon S3
		Creating buckets to hold data
		Distinguishing between object tags and object metadata
	Controlling access to buckets and objects on amazon s3
		S3 bucket policy
	Protecting data on amazon s3
		Applying bucket versioning
		Applying encryption to buckets
	Securing s3 objects at rest and in transit
	Using other types of data stores
	Relational Database Service (RDS)
	Managing failover in Amazon RDS
	Taking automatic backups, RDS snapshots, and restore and read replicas
	Writing to Amazon Aurora with multi-master capabilities
	Storing columnar data on Amazon Redshift
	Amazon DynamoDB for NoSQL Database-as-a-Service
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 3: AWS Services for Data Migration and Processing
	Technical requirements
	Creating ETL jobs on AWS Glue
		Features of AWS Glue
		Getting hands-on with AWS Glue Data Catalog components
		Getting hands-on with AWS Glue ETL components
	Querying S3 data using Athena
	Processing real-time data using Kinesis Data Streams
	Storing and transforming real-time data using Kinesis Data Firehose
	Different ways of ingesting data from on-premises into AWS
		AWS Storage Gateway
		Snowball, Snowball Edge, and Snowmobile
		AWS DataSync
		AWS Database Migration Service
	Processing stored data on AWS
		AWS EMR
		AWS Batch
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 4: Data Preparation and Transformation
	Identifying types of features
	Dealing with categorical features
		Transforming nominal features
		Applying binary encoding
		Transforming ordinal features
		Avoiding confusion in our train and test datasets
	Dealing with numerical features
		Data normalization
		Data standardization
		Applying binning and discretization
		Applying other types of numerical transformations
	Understanding data distributions
	Handling missing values
	Dealing with outliers
	Dealing with unbalanced datasets
	Dealing with text data
		Bag of words
		TF-IDF
		Word embedding
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 5: Data Understanding and Visualization
	Visualizing relationships in your data
	Visualizing comparisons in your data
	Visualizing distributions in your data
	Visualizing compositions in your data
	Building key performance indicators
	Introducing QuickSight
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 6: Applying Machine Learning Algorithms
	Introducing this chapter
	Storing the training data
	A word about ensemble models
	Supervised learning
		Working with regression models
			Introducing regression algorithms
			Least squares method
			Creating a linear regression model from scratch
			Interpreting regression models
			Checking adjusted R squared
			Regression modeling on AWS
		Working with classification models
		Forecasting models
			Checking the stationarity of time series
			Exploring, exploring, and exploring
			Understanding DeepAR
		Object2Vec
	Unsupervised learning
		Clustering
			Computing K-Means step by step
			Defining the number of clusters and measuring cluster quality
			Conclusion
		Anomaly detection
		Dimensionality reduction
			Using AWS’s built-in algorithm for PCA
		IP Insights
	Textual analysis
		BlazingText algorithm
		Sequence-to-sequence algorithm
		Neural Topic Model algorithm
	Image processing
		Image classification algorithm
		Semantic segmentation algorithm
		Object detection algorithm
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 7: Evaluating and Optimizing Models
	Introducing model evaluation
	Evaluating classification models
		Extracting metrics from a confusion matrix
		Summarizing precision and recall
	Evaluating regression models
		Exploring other regression metrics
	Model optimization
		Grid search
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 8: AWS Application Services for AI/ML
	Technical requirements
	Analyzing images and videos with Amazon Rekognition
		Exploring the benefits of Amazon Rekognition
		Getting hands-on with Amazon Rekognition
	Text to speech with Amazon Polly
		Exploring the benefits of Amazon Polly
		Getting hands-on with Amazon Polly
	Speech to text with Amazon Transcribe
		Exploring the benefits of Amazon Transcribe
		Getting hands-on with Amazon Transcribe
	Implementing natural language processing with Amazon Comprehend
		Exploring the benefits of Amazon Comprehend
		Getting hands-on with Amazon Comprehend
	Translating documents with Amazon Translate
		Exploring the benefits of Amazon Translate
		Getting hands-on with Amazon Translate
	Extracting text from documents with Amazon Textract
		Exploring the benefits of Amazon Textract
		Getting hands-on with Amazon Textract
	Creating chatbots on Amazon Lex
		Exploring the benefits of Amazon Lex
		Getting hands-on with Amazon Lex
	Amazon Forecast
		Exploring the benefits of Amazon Forecast
		Sales Forecasting Model with Amazon Forecast
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 9: Amazon SageMaker Modeling
	Technical requirements
	Creating notebooks in Amazon SageMaker
		What is Amazon SageMaker?
		Training Data Location and Formats
		Getting hands-on with Amazon SageMaker notebook instances
		Getting hands-on with Amazon SageMaker’s training and inference instances
	Model tuning
		Tracking your training jobs and selecting the best model
	Choosing instance types in Amazon SageMaker
		Choosing the right instance type for a training job
		Choosing the right instance type for an inference job
	Taking care of Scalability Configurations
		Scaling Policy Overview
		Scale Based on a Schedule
		Minimum and Maximum Scaling Limits
		Cooldown Period
	Securing SageMaker notebooks
	SageMaker Debugger
	SageMaker Autopilot
	SageMaker Model Monitor
	SageMaker Training Compiler
	SageMaker Data Wrangler
	SageMaker Feature Store
	SageMaker Edge Manager
	SageMaker Canvas
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 10: Model Deployment
	Factors influencing model deployment options
	SageMaker deployment options
		Real-time endpoint deployment
			Solution
			Steps
			Example code snippet
		Batch transform job
			Solution
			Steps
			Example code snippet
		Multi-model endpoint deployment
			Solution
			Steps
			Example code snippet
		Endpoint autoscaling
			Solution
			Steps
			Example code snippet
		Serverless APIs with AWS Lambda and SageMaker
			Solution
			Steps
			Example code snippet
	Creating alternative pipelines with Lambda Functions
		Creating and configuring a Lambda Function
		Completing your configurations and deploying a Lambda function
	Working with step functions
	Scaling applications with SageMaker deployment and AWS Autoscaling
		Scenario 1 – Fluctuating inference workloads
			Autoscaling solution
			Steps
			Example code snippet
		Scenario 2 – The batch processing of large datasets
			Autoscaling solution
			Steps
			Example code snippet
		Scenario 3 – A multi-model endpoint with dynamic traffic
			Autoscaling solution
			Steps
			Example code snippet
		Scenario 4 – Continuous Model Monitoring with drift detection
			Autoscaling solution
			Steps
	Securing SageMaker applications
	Summary
	Exam Readiness Drill – Chapter Review Questions
Chapter 11: Accessing the Online Practice Resources
Index
Other Books You May Enjoy