Machine Learnng

Question 1

Amazon Rekognition

Answer 1

• graphic processing - finds objects, text,scenes, images, poeple, videos using ML
• face detecion and search and analysis
• creates database of familiar faces, compare
  Use case:
• labeling
• text detection
• celebrity reckognition

Amazon Rekognition – Content Moderation -
* detects content that is** inappropiate or offensive**
* you can set minimum confidence treshold
* flag sensitive content for manual review

Question 2

Amazon Transcribe

Answer 2

• speech to text
• ASR (Automatic speech reckognition)
• automatic language detection
• automatic detection of PII
• transacribe customer service calls

Question 3

Amazon Polly

Answer 3

• text into speech
• Pronunciation lexicons => acronyms, slylized words - use in SynthesizeSpeech
• SSML - emphasizing specific words, whispering, phonetic pronounciation

Question 4

Amazon Translate

Answer 4

• language translation

Question 5

Amazon Lex

Answer 5

• speech to text
• natural language understanding
• chatbots Amazon Lex enables any developer to build conversational chatbots quickly

Question 6

Amazon Connect

Answer 6

• receive calls
• create flow

Question 7

Amazon Comprehend

Answer 7

• NLP natural language processing
• uses ML to analyse text and find nsights
• organise text files bytopic
• analyze email history to understand leads

Question 8

Amazon Comprehend Medical

Answer 8

• Amazon Comprehend Medical detects and returns useful information in
  unstructured clinical text:
• Physician’s notes
• Discharge summaries
• Test results
• Case notes
• detects Protected Health Information PHI
  You can use the extracted medical information and their relationships to build applications for use cases, like clinical decision support, revenue cycle management (medical coding), and clinical trial management. Because Comprehend Medical is HIPAA-eligible and can quickly identify protected health information (PHI), such as name, age, and medical record number, you can also use it to create applications that securely process, maintain, and transmit PHI.

Question 9

Amazon SageMaker

Answer 9

• build ML models in fast and easy way

Question 10

Amazon Forecast

Answer 10

• delivers** forecast**
• example: predict future sales of jeans
• very quick
• planning: financial, resource etc

Question 11

Amazon Kendra

Answer 11

* document search
* extract text from documant
* IncrementalLearning learn from user interactions feedback to promote preferred results
* able to filter search by importance of data, freshness etc.
* user asks question and kendra gicen back result based on knowledge from given documents

Question 12

Amazon Personalize

Answer 12

• similar to Amazon - marketing technology tool
• build app with real-time personalized recommendations
• ready ML solution no data training

Question 13

Amazon Textract

Answer 13

• automatic extraction of handwriting from files and documents