AWS Series — Machine Learning tools in AWS that you might have to know
In AWS, the number of tools and services are expanding so as the Machine learning tools. Let us look at high level what they are and what they do.
Analyzing Text using Amazon Comprehend, Amazon Kendra and Amazon Textract
Amazon Comprehend
- Comprehend Uses Natural Language Processing to help you understand the meaning and sentiment in your text.
- You can also pick up on key phrases. It is a way of automating comprehension at scale.
Amazon Kendra
Amazon Textract
Predicting Time Series Data using Amazon Forecast
Protecting Accounts with Amazon Fraud Detector
Use cases for using this service
- Identify Suspicious Online Payments
- Detect new Account Fraud
- Prevent Trial and Loyalty Program Abuse
- Improve Account Takeover detection
Working with Text And Speech using Amazon Polly, Amazon Transcribe and Amazon Lex
Alexa which you might definitely know, uses Amazon Transcribe to convert speech to text that we talk and send to Lex service which creates bot response and it is responded using Polly service that talks like human with some accent.
Amazon Rekognition
Usecases for Rekognition
- Content Moderation — Automatically moderate content allowing your applications and websites to be family friendly
- Celebrity Recognition — Automatically recognizes celebrity people and label them
- Face Detection and Analysis — Automatically recognizes faces and detect whether someone is wearing a hat or glasses
- Streaming Video Event Detection — Useful for applications like Ring which recognizes the person and image and automatically creates alerts
Amazon SageMaker
Sage Maker Neo
Customize your machine learning models for specific CPU hardware such as ARM, Intel, and NVIDIA processors
It includes a compiler to convert the machine learning model to an environment that is optimized to execute the model on the target architecture.
Elastic Inference
- Speeds up throughput and decreases latency of real-time inferences deployed on SageMaker hosted services using only CPU-based instances. It is much more cost effective than a full GPU instance.
- It must be configured when you create a deployable model. EI is not available for all algorithms yet.
Sagemaker comes with Automatic Scaling. It is also recommended to deploy in multiple Availability Zone for making it highly available.
Translating Content into Different Languages with Amazon Translate
- ML service that allows to automate language translation.
Happy Learning!!
