Download to read offline
Uploaded byMeetupDataScienceRoma
Machine Learning in the AWS Cloud
The document provides use cases and solutions for building various machine learning applications using Amazon Web Services. It discusses how to create a speech enabled facial recognition system using Amazon Rekognition and Amazon Polly. It also discusses how to build a chat app with sentiment analysis using Amazon Comprehend, Amazon Lex, and Amazon Translate. Additional use cases discussed include podcast episode discovery and indexing using Amazon Transcribe and Amazon Comprehend, and building a recommendation system using Amazon SageMaker.
More Related Content
Similar to Machine Learning in the AWS Cloud
![[Giovanni Galloro] How to use machine learning on Google Cloud Platform](https://cdn.slidesharecdn.com/ss_thumbnails/mlcapabilitiesongcp-190115085455-thumbnail.jpg?width=640&height=640&fit=bounds)
![[Sponsored] C3.ai description](https://cdn.slidesharecdn.com/ss_thumbnails/c3deckmeetupnovember252019v2-191128092146-thumbnail.jpg?width=640&height=640&fit=bounds)
![Support, Monitoring, Continuous Improvement & Scaling Agentic Automation [3/3]](https://cdn.slidesharecdn.com/ss_thumbnails/agenticcommunityseries-day3-cfd-251120170304-ddef8112-thumbnail.jpg?width=640&height=640&fit=bounds)
Machine Learning in the AWS Cloud
- 2.
- 3. Reasons to ReuseNot Reinvent Retain Functionality Increase EfficiencyRetain Investment
- 4. Amazon Web Services APPLICATIONSERVICE PLATFORMS FRAMEWORKS / INTERFACES
- 5. 01How to createa speech enabled system for facial recognition? Real Use Case
- 6. Use Case The aimis to create a system for facial recognition that, once the person has been recognised, will also welcome him/her. The application must be scalable and serverless. Facial Recognition Speech-enabled Application Video Source
- 7. How can yourapplications see the world? Amazon Rekognition Identifies objects, people, text, scenes, faces and activities, as well as detects any inappropriate content inside an image or video. Object & Scene Detection Facial Search & Analysis Celebrity Recognition
- 8. Amazon Polly Text-to-speech servicebased on advanced deep learning technologies to synthesize speech that sounds like a human voice. How can your applications have a voice?
- 9. 02 How to createa chat app with Sentiment Analysis? Real Use Case
- 10. Use Case PubNub isa leading provider of real-time APIs for building chat, device control and real-time mapping apps that scale globally. PubNub ChatEngine has integrated Amazon machine learning APIs. Cross-lingual Application Conversational Interfaces Sentimental Analysis
- 11. How to extractinsights from text? Amazon Comprehend A fully managed and continuously trained service that helps you extract insights from unstructured text Sentiment Key PhrasesEntities Languages Topic Modelling
- 12. How do youmake your app conversational? Amazon Lex A service for build conversional interfaces into your applications using voice and text
- 13. How do youmake your app multilingual? Amazon Translate A fully managed and continuously trained neural machine translation service that translates text from one language to another 12 Languages & more to come Translate Text Input Real-time Translation
- 14.
- 15. 03 How to perform discoveringand indexing of podcast episodes?
- 16. Use Case Build atool that converts the audio to text and then build a searchable index of podcast feeds to discover information without having to listen to a full episode. Not all episode abstracts are equally helpful! Audio Transcribing Text Comprehend Text Indexing
- 17. How do youmake your app listen? Amazon Transcribe A fully managed and continuously trained automatic speech recognition (ASR) service that takes in audio and automatically generates accurate transcripts Regular & Telephony Amazon S3 Integration Time Stamps & Confidence Scores Punctuation Detect multiple speakers Custom vocabulary
- 18.
- 19. 04How to builda recommendation system? Real Use Case
- 20. Use Case Condé NastInc. is an American mass media company, it attracts more than 164 million consumers across its brands: VanityFair, Vogue, GQ, etc. It needs a recommendation system to improve customer experience. Data Transformation Model Training Model Deploy
- 21. Amazon SageMaker (1) Itis a fully managed service that provides the quickest and easiest way for your data scientists and developers to build, train and deploy Machine Learning models… ..from idea to production.
- 22.
- 23. Machine Learning LifeCycle Business Problem Re-training Predictions No Yes DataAugmentation Feature Augmentation Are Business Goals met?
- 24. Condé Nast, Hybrid.Data Processing • Remove! Most frequent words • Remove! Punctuation, symbols, numbers • Keep! adjectives, verbs, adverbs, pronouns • Lemming! remove inflectional and to return the base or dictionary form of a word: said, say.. • NGram! Words that have different meanings if together "New York" WORDS = 0.579 STOP WORDS = 0.421
- 25. Condé Nast, Hybrid.Training Documents Similarity Matrix !", "$ = &. ( "$, !" = &. ( !", !" = !
- 26. Condé Nast, Hybrid.Recommend !"#"$%&"'( 2, 55 = 70% → 1234##256 %&"'55 7$"389 :92& = %&"'0, %&"'1, %&"'2 <%'&"= = >='&%3' <%'&"=, 7$"389 =
- 27. Condé Nast, Hybrid.Evaluation After training no one can say if it's okay or not okay because there could be another Dataset for which you get different results, but we must first find out! • Test Dataset • Aphorisms • Download sentences with labels • 34 Classes • Love • Friendship • Woman • Man • etc • Complex • Ambiguous 5% 40%
- 28. Amazon SageMaker (&3) EasyModel Deployment to Amazon SageMaker Hosting
- 29.
- 30.