AWS Machine Learning (15 ECTS)
Code: TT00FA94-3001
General information
- Timing
- 01.08.2022 - 31.12.2023
- Implementation has ended.
- Number of ECTS credits allocated
- 15 ECTS
- Virtual portion
- 15 ECTS
- Mode of delivery
- Online
- Unit
- (2019-2024) School of ICT
- Campus
- Karaportti 2
- Teaching languages
- English
- Seats
- 0 - 500
- Degree programmes
- Information and Communication Technology
- Teachers
- Virve Prami
- Groups
-
OPEN_UAS_TIVI_AI_ML_DS_75_ECTSOpen UAS: Artificial intelligence, Machine Learning and Data Science (NonStop Module) 75 ECTS
- Course
- TT00FA94
Objective
AWS Machine Learning services provide anybody with access to the same cloud-based machine learning resources used by Amazon's own developers for machine learning implementations. It offers developers and data scientists various resources to help build their knowledge of machine learning in the AWS Cloud. This course will teach you how to get started with AWS Machine Learning and focuses on standardized approaches to data analytics and machine learning implementation based on AWS Machine Learning tools. By the end of this course, the student will learn to build, train, deploy, automate, manage, and track enterprise-grade machine learning models from scratch in a simplified way using powerful AWS tools.
This course is 100% virtual thanks to the comprehensive tutorial videos and content prepared for this course.
The student will pass this course after submitting the required quiz, assignments, and the final project.
Content
1. Introduction:
Basics of Machine Learning – Importance of Data – Types of Machine Learning – Machine Learning Pipeline – Machine Learning Implementation – Why Cloud-based Services? – Amazon Web Services (AWS)
2. AWS Machine Learning:
What is AWS Machine Learning Service? – AWS AI Services – Amazon SageMaker – SageMaker Studio – SageMaker Autopilot – SageMaker Studio Notebooks – SageMaker Pipelines – Quiz
3. Getting Started with AWS:
AWS Pricing – Compute Instance Types – AWS Free Tier Account – Create a Free Tier Account – A Tour of AWS Interface – IAM Service – Getting Started with AWS: Exercise
4. Amazon S3:
Introduction to S3 – Set Up Amazon S3 – Create a Bucket –, Download, and Copy an Object – Properties for an S3 Bucket – Bucket Permissions – Accessing a Bucket – Storage Tier and Data Lifecycle – Delete an Object or a Bucket – Amazon S3: Exercise – Amazon S3: Quiz
5. Amazon SageMaker Autopilot:
Set up Amazon SageMaker Studio – Getting Started with SageMaker Autopilot – Create a SageMaker Autopilot Experiment – Problem Types – Model Support and Validation – Model Deployment – Models Generated by SageMaker Autopilot – Notebooks Generated by SageMaker Autopilot – Configure Inference Output – SageMaker Autopilot Quotas
6. Amazon SageMaker Autopilot: Practical Example:
Open SageMaker Studio – Load the Dataset – Create an Experiment – Explore the Experiment – Data Exploration Notebook – Deploy the Best Model – Predict with Your Model – Job Profile – Trials: Deploy the Best Model – Inference on SageMaker Console: Endpoints – on SageMaker Console: Training Job
6.12. Training on SageMaker Console: Hyperparameter Tuning Jobs – Delete Endpoints Manually – Project 1
7. Amazon SageMaker Studio Notebooks:
Studio Notebooks vs. Instance-based Notebooks – Create a Notebook in Studio – Access Notebook Instances – What is SageMaker Python SDK? – Prepare a Training Script – Upload and Insert Libraries – Upload Data on Jupyter Instance – Upload Data on S3 Buckets – Load Data – Data Preprocessing – Model Building Process – Model Building Example – Deploy the Model – Evaluate Model Performance – Terminate the Resources – Project 2
8. SageMaker Pipelines:
Introduction – Machine Learning Workflow Challenges – Amazon SageMaker Pipelines – SageMaker Pipelines Components – Create a Project – Explore the Repository – Processing Step – Training Step – Evaluation Step – Pipeline Execution – Register the Model to Registry – Approve the Model for Deployment – Project 3
9. Final Tasks
Self-study Essay – Congrats! You did it!
Location and time
Course can be done in own pace in TechClass portal.
Materials
Lecture slides, tutorial videos, quizzes, exercises and project can be find in study environment.
Teaching methods
This course is 100% virtual thanks to the comprehensive tutorial videos and content prepared for this course.
The student will pass this course after submitting the required quiz, assignments, and the final project.
Employer connections
N/A
Exam schedules
Online.
International connections
N/A
Completion alternatives
N/A
Student workload
- Tutorial Videos
- Exercises
- Quiz
- Project
- Self-study
Evaluation scale
Hyväksytty/Hylätty
Assessment criteria, satisfactory (1)
Evaluation criteria - satisfactory (1-2)
- The student is familiar with cloud-based services.
- The student is familiar with different services that AWS offers for AI and Machine Learning.
- The student is familiar with Amazon SageMaker.
- The student is familiar with the concept of Automated Machine Learning (AutoML) as well as the SageMaker Autopilot for using AutoML capabilities.
- The student is familiar with SageMaker Studio, how to create Notebooks in Studio, and how to use the basic functionalities of the Studio.
- The student is familiar with creating a free-tier AWS account and setting up compute instances.
- The student is familiar with Amazon S3 and knows how to use the primary functionalities of S3.
- The student knows how to use the basic functionalities of SageMaker Autopilot to train Machine Learning models.
Assessment criteria, good (3)
Evaluation criteria - good (3-4)
- The student knows how to use SageMaker Autopilot to load data and create and configure experiments.
- The student knows how to use SageMaker Autopilot to train models and deploy them to the AWS cloud.
- The student knows how to use SageMaker Autopilot and job profiles to explore trained models, choose the best model, get inference from the model, and deploy it to the cloud.
- The student knows how to use SageMaker Autopilot for tuning models’ hyperparameters.
- The student knows the difference between SageMaker Studio notebooks and SageMaker Instance-based notebooks.
- The student knows how to create Jupyter notebooks in SageMaker, attach EC2 compute instances to them, and run them on the cloud
- The student knows how to connect to S3 from SageMaker notebooks.
- The student is familiar with SageMaker Pipelines, its components, and use cases.
- The student knows how to create simple pipelines using SageMaker Pipelines.
Assessment criteria, excellent (5)
- The knows how to use SageMaker Autopilot to do build and train supervised Machine Learning models with advanced hyperparameter tuning and performance evaluation
- The knows how to use SageMaker Autopilot to do inference on SageMaker Console using endpoints
- The student knows how to use SageMaker Python SDK to prepare training scripts, upload and insert libraries, and upload data on Jupyter instance.
- The student knows how to use SageMaker Python SDK to load different datasets and perform data preprocessing and preparation.
- The student knows how to use SageMaker Python SDK to build supervised Machine Learning models, train the models on training sets, deploy them to EC2 compute instances, and evaluate their performance.
- The student knows how to use SageMaker Pipelines to create pipelines for processing data using different AWS services.
- The student knows how to use SageMaker Pipelines to create pipelines with Machine Learning model training functionalities.
- The student knows how to use SageMaker Pipelines for creating automatic pipelines for training, evaluating, and deploying Machine Learning and using them for prediction
Assessment methods and criteria
Exercises 50%
Quizzes 25%
Project 25%
Qualifications
Fundamentals of Machine Learning or Machine Learning with python