This course will be delivered using the following platforms

Note - This course is presented entirely free of cost to participants.

SDCCD Canvas: Our communications hub, where we will find our weekly schedule and announcements, links to our Online Live Sessions in SDCCD Zoom, a place to submit our lab assignments, and our course grade book. Use your student CSID and password to log in.

AWS Academy hosted on Canvas LMS: This is where we will find our e-learning resources, including pre-recorded video lessons and demonstrations, as well as our graded assignments. We will access this from a web browser (Chrome or Firefox) on a standard PC or laptop (Windows, Mac, or Linux - no tablets or smartphones). This is not affiliated with SDCCD and your student account - to log-in, we will be provided with a unique AWS Academy Canvas student account.

You will be provided with separate access to the AWS Academy LMS by your instructor. Expect to receive an email from Instructure at the start of the course inviting you to complete your AWS Academy student account and join the course.

AWS Management Console: This web application is comprised of a broad collection of service consoles for managing AWS resources. This, along with the AWS CLI and SDK's is one of the three primary methods AWS cloud practitioners use to plan, deploy and manage their cloud-based solutions. We will access this, free of charge, from a web browser on the AWS Academy in Canvas, to complete our curated set of hands-on labs.

Zoom Video Conferencing: To attend the weekly Online Live Sessions, we will be using Zoom through our college at SDCCD. The meeting registration link is provided on the course home page in SDCCD Canvas. For the best experience, use a computer or tablet with a minimum screen size of 10 inches that includes a speaker and microphone. We also recommend a video camera. You can use a mobile phone in a pinch. If you are new to Zoom, check out these "What is Zoom Video Conferencing" or these short videos to learn about "Joining a Zoom Meeting" and "Basic In-Meeting Navigation". If you experience trouble joining a meeting, check out the Zoom Help Center.

Course Topics

• WEEK ONE Course Introduction (AWS Mod-1)

• WEEK TWO Introducing Machine Learning (AWS Mod-2)

• WEEK THREE Machine Learning Pipeline Part 1 (AWS Mod-3 Sections 1-3)

• WEEK FOUR Machine Learning Pipeline Part 2 (AWS Mod-3 Sections 4-6)

• WEEK FIVE Machine Learning Pipeline Part 3 (AWS Mod-3 Sections 7-8)

• WEEK SIX  Introducing Forecasting (AWS Mod-4)

• WEEK SEVEN Introducing Computer Vision (AWS Mod-5)

• WEEK EIGHT Introducing Natural Language Processing (AWS Mod-6)

• WEEK NINE Introducing Generative AI & Wrap Up (AWS Mod-7 & 8)

Course Goal

The goal of this course is to provide practical experience in the selection and application of AWS managed services and tools for common machine learning tasks.

Course Objectives

NOTE: Each one of the following course objectives is covered in single module. For a detailed description of each objective and their accompanying module objectives, see the TOPIC SCHEDULE.

After completing the ten modules in this course, you should be able to:

• CO 1: Describe machine learning.

• CO 2: Implement a machine learning pipeline using Amazon SageMaker.

• CO 3: Use managed Amazon machine learning services for forecasting.

• CO 4: Use managed Amazon machine learning services for computer vision.

• CO 5: Use managed Amazon machine learning services for natural language processing.

• CO 6: Identify how Amazon machine learning services for generative AI are used.

SDCCE Institutional Learning Outcomes

1. Social Responsibility SDCCE students demonstrate interpersonal skills by learning and working cooperatively in a diverse environment.

2. Effective Communication SDCCE students demonstrate effective communication skills.

3. Critical Thinking SDCCE students critically process information, make decisions, and solve problems independently or cooperatively.

4. Personal and Professional Development SDCCE students pursue short term and life-long learning goals, mastering necessary skills and using resource management and self advocacy skills to cope with changing situations in their lives.

5. Diversity, Equity, Inclusion, Anti-racism and Access SDCCE students critically and ethically engage with local and global issues using principles of equity, civility, and compassion as they apply their knowledge and skills: exhibiting awareness, appreciation, respect, and advocacy for diverse individuals, groups, and cultures.

Business Information Technology (BIT)

Mission Statement: The Business and Information Technology Program (BIT) provides adults open access to transformational career technical education programs. Through skill building, upskilling and reskilling, BIT provides the San Diego community the opportunity to transition to college and work by providing hands-on and project based training in current technology, foundational skills, and business practices with real-work simulations. Lead by industry experienced instructors; these programs build student confidence for future employment, promotion, and entrepreneurial opportunities.

BIT Department SLO's

1. Students completing a BIT software course will be able to demonstrate the use of the software tools to effectively communicate with others in person, with paper documents or online. (Relates to Institutional SLO #2 above)

2. BIT students work in teams with diverse individuals to apply Information Technology solutions to a problem. (Relates to Institutional SLO #1 above)

3. BIT students use Information Technology and software tools to support decision processes and critical thinking. (Relates to Institutional SLO #3 above)

4. BIT students pursue continued Information Technology education to complete short term goals such as website development, and also continue with long term programs that will keep them current in this rapidly changing field. (Relates to Institutional SLO #4 above)

Completion Requirements

Achieve a "C" grade or higher by accomplishing the following:

• Accumulate a minimum of 910 points total (70%)

Assignment Policies

• Due Dates: Assignments should be submitted the week in which the Academy module is covered.

• Resubmission: Assignments may be submitted multiple times until the course end date (or the lab wallet is expended) without a grade penalty.

• Late: Assignments may be submitted after the weekly due date posted on the course schedule without a grade penalty.

Grade Scale

• A = 100.00 to 95.00% = 1,300 to 1,235 points

• B =  94.99 to 85.00% = 1,234 to 1,105 points

• C =  84.99 to 70.00% = 1,104 to 910 points

• D =  69.99 to 60.00% = 909 to 780 points

• F =  59.99 to 00.00% = 779 to 0 points

Types of Assignments and Activities

• AWS Academy Knowledge Checks (KC) - 100 points each, ten questions each, six total for up to 600 points
  Each AWS Academy module includes a multiple-choice formative assessment. The assignment is intended to be used as a learning tool and provides immediate response feedback, with multiple attempts available as needed to allow us to gain greater confidence with the material and improve our scores. Note that the grade book displays the most recent score. 

• AWS Hands-on Labs (HOL) - 100 points each, seven total for up to 700 points
  Hands-on labs are presented within the AWS Academy modules, and are accessible using the Vocareum lab platform. These HOL's provide access to AWS services using the AWS Management Console, a web application accessible through a web browser (such as Chrome and Firefox) for managing AWS resources. This allows us to experience the AWS Cloud services first-hand using the same tools used by professionals. Upon completion of each lab, save evidence of completion by recording the required deliverable and submit it to the SDCCD Canvas assignment submission page. Note that the grade book maintains your most recent score. Here is a list of our hands-on labs. For lab descriptions, see the Hands-on Labs section below. 

Note: Two labs are excluded from the grade calculation, and are noted by a pair of asterisks

• • • Module 3 ML Pipeline on Sagemaker (500 points):

      • Lab 3.1 - Amazon SageMaker - Creating and Importing Data - 100 points 

      • ** Lab 3.2 - Amazon SageMaker - Exploring Data (no submission)

      • Lab 3.3 - Amazon SageMaker - Encoding Categorical Data (100 points)

      • ** Lab 3.4 - Amazon SageMaker - Training a Model (no submission)

      • Lab 3.5 - Amazon SageMaker - Deploying a Model (100 points)

      • Lab 3.6 - Amazon SageMaker - Generating Model Performance Metrics (100 points)

      • Lab 3.7 - Amazon SageMaker - Hyperparameter Tuning (100 points)

    • Lab 5 - Guided Lab: Facial Recognition (100 points)

    • Lab 6 - Amazon Lex - Create a Chatbot (100 points)

Submitting Assignments

Complete assignments weekly according to the schedule. Minimum one graded assignment (KC or HOL) each week to meet minimum participation requirements.

Submitting assignments

• AWS Knowledge checks are completed in the AWS Academy.

  • The instructor will update the SDCCD grade book with your most current score.

• AWS Hands-on labs are completed in the AWS Academy.

  • Before attempting the assignment, check the SDCCD assignment submission page for instructions.

  • You must submit evidence of completion on the SDCCD assignment submission page to earn points for the lab.

  • Failure to submit the required evidence will result in zero points for the assignment/

  • Upon ending a lab session, all resources launched or deployed (either by Vocareum at the start of the lab or by you in the course of the session) will be terminated and removed.

• Assignments may be attempted and submitted multiple times, with the most recent score displayed in the course gradebook

• All assignments are due by the last day of the course. After that date, they will not be included in the calculation of your final grade

Late and Missing Work

I want to help you achieve your goals in this course - if you are struggling to meet the schedule, contact me as soon as possible so we can discuss this and create a plan that will help you succeed!

Completing this course along with its companion, provides eligibility for several useful badges, vouchers, and certificates.

AWS Academy Course Digital Badge

The minimum AWS Academy requirement is completion of all knowledge checks with a minimum score of 70% each. You will receive an email within 24 hours from Amazon Web Services Training and Certification via Credly to claim your digital badge and downloadable certificate. You can then share our badge on our LinkedIn or other social media profile to let peers and potential employers know about our accomplishment. Instructions for receiving the course badge may be found in the AWS Academy on Canvas modules page.

SDCCD Cloud Solutions AWS-Specialty Machine Learning Program Certificate

Upon successfully completing the three courses in the certificate program, you will be awarded the SDCCD program certificate:

• COMP 671 AWS Academy Cloud Foundations

• COMP 672 AWS Academy Cloud Architecture

• COMP 675 Academy Machine Learning Foundations

You can view the award and download an unofficial transcript in your student portal at myportal.sdccd.edu.

I want and know that you can succeed in this course, and I have found that regular weekly participation is one of the most effective ways to learn and grow your Cloud skills. To help make that happen, this course is offered online and synchronous, which means that we will have regular weekly online meetings and activities.

So what is regular participation?

Regular participation means check into the course at minimum three times a week:

• Completing two assignments each week, should include at least one hands-on-lab and a second graded assignment

• Attending Weekly Live Sessions

• Viewing module videos and reading

• Practicing demos with the sandbox

• Responding to messages from the instructor within 48 hours, or sooner if urgent

Note: If you miss an Online Live Session or I do not hear from you and you do not participate in the course for over a week, I will send you a Canvas message. If I do not hear back from you within 24 to 48 hours, and you still have not accessed the course, I may assume you have dropped, and will remove your name from the course roster.

I have designed the course to provide us with a mix of both concept and practical hands-on experience, which together helps us to grow our understanding of cloud computing. It is to everyone’s benefit that we do our best to complete assignments weekly. However, I also recognize that situations can occur that may prevent you from meeting this schedule. In general, do your best to stay current with the weekly material. If you cannot participate regularly or know that you may have to miss a week in Canvas for an unavoidable circumstance, let me know right way. Stay in contact and respond to any messages within 48 hours.

Our meetings are held weekly using the Zoom video conferencing platform

• Meeting Registration: Link is available in SDCCD Canvas on the SDCCD Zoom page

• Purpose: Interact with the instructor & classmates in discussions, group activities, and ask and answer questions

• Activities: Combination of case studies and scenarios, and demos.

• Sample exam questions: Learn how to analyze and answer certification exam questions

• Live demonstrations: Using the AWS management console and AWS CLI

If you miss a Live Session:

• Contact the instructor the same day, or in advance.

• View the meeting recording (available from the course home page in SDCCD Canvas only

• Check announcements and email to ensure you are up-to-date on any changes or important messages

I value your success and I know your ability to communicate with me is an important ingredient in that recipe.

• Contact me Monday through Friday by Canvas Inbox, and I will respond within 24 to 48 Hours.

• Meet with me in Zoom before or after the weekly Live Session.

• Meet with me in Zoom during Student Virtual Office Hours.

• If you are seeking help with a lab, consider scheduling time in Zoom to work on it together!

Canvas Inbox: It is important to stay in contact, and this is one of the best ways to do so. I will respond to your message within 48 hours (but usually sooner), Monday – Friday before 5:30 PM. You can either check your messages in the CANVAS system or set your notifications to your preferred method of contact. If you send me a message over the weekend or during the holiday, expect a response by Monday or Tuesday afternoon.

Canvas Announcements: You will receive one each week on Sunday when the weekly module opens. These appear at the top of the class homepage when you log in and will be sent to you directly through your preferred method of notification from CANVAS. Check them regularly, as they contain important information about upcoming assignments or class concerns.

If I do not hear from you, and your course participation drops, I will reach out through Canvas Inbox, to make sure everything is alright. It is important that you respond as soon as you receive the message. Remaining in communication with myself (and your classmates) is one of the best ways to ensure success in the course.

Help with Lab Assignments: If you are seeking help with an assignment, include the assignment name and number, the specific step number, and any error messages and relevant information, including the expected outcome. The more accurate and specific, the better. Sometimes a screen shot or two can explain things that words cannot, especially when properly annotated. You might also consider dropping by the weekly office hours in Zoom or during the Live Session, or we can schedule a one-to-one Zoom session.

Counseling/Student Services

Student services provides If you need help with a personal problem or advice about your studies, you can make an appointment with a counselor.  For example, a counselor can help you make a plan to reach your goals: improving your English, getting your GED, enrolling in a job training class or attending college.  If you need help finding a job, you can contact the Career Development Services Counselor

• Course Counselor: Joyce Almario-Greno, jalmario@sdccd.edu

• Job Developer: Jennifer Kennedy, 619-800-3093, jkennedy@sdccd.edu

• Contact Career Services

Disability Support Programs and Services (DSPS)

If you have a disability or think you might have a disability, you can contact the counselor in the Disability Support Programs and Services (DSPS) at your campus. DSPS can provide services and special equipment that will make it easier for you to study in our classes. An example of special equipment is a machine that enlarges the print for people who have a vision disability. Since it takes time to provide services, we recommend that you contact the counselor at least two weeks in advance.  DSPS services are confidential and voluntary.

Enrollment Assistance

For assistance with your SDCCD student password or student records: Use the secure mySDCCD Support Desk. Complete the top portion, and at the bottom of the web page, select from the Help Topic "I forgot my password". You will then be required to submit a digital copy of your government issued ID for proof of identity.

To Speak with Live Staff: Sign up for our Virtual Student Support Center (Links to an external site.)

For all other matters: email the campus at sdcenorthcity@sdccd.edu or sdcemesa@sdccd.edu. All of the staff are waiting to help students.

Course Attendance

PARTICIPATION REQUIREMENTS

To maintain active status in the course, regular attendance is expected:

• Submit at least one AWS assignment each week

• Regularly attend our Live Sessions

• Respond to messages within 48 hours

• Be proactive and contact the instructor if you are not able to meet these expectations

Plan to check into the course at minimum 3 times a week. Any student frequently absent from the course may, at the discretion of the instructor, be dropped from the course. Those students receiving Veteran’s Benefits or CalWORKS must comply with the attendance requirements specific to these programs.

BP 5500 - Student Rights, Responsibilities, Campus Safety & Administrative Due Process - This policy enumerates the rights and responsibilities of all District students. It also outlines the District’s commitment to a safe learning environment for all students.

Commitment to Integrity and Academic Dishonesty Policy

• Students should actively participate in course activities.

• Our college has rules about academic dishonesty:

  • Students are not permitted to cheat on course assignments or tests.

  • Students are not permitted to use false information.

  • Students may not copy the language or ideas of another person and use them as their own ideas.

• An instructor will take the following steps if he/she thinks a student has been dishonest in completing a course assignment or test:

  • Discuss the situation with the student. Make sure that the student understands why his/her action is dishonest.

  • If the student did not understand that his action was dishonest, the instructor can give the student a warning.

  • If the student knew that his action was dishonest, the instructor can give him/her a failing grade.

Note that live sessions fall on the day of the week and at the times provided to you before the term start and proceed in a weekly manner. Live sessions will not be held on SDCCE holidays.  If a live session for this course falls on an SDCCE holiday, the live session will be rescheduled, and your instructor will inform you as to when the Live Session will be rescheduled or how the content will be covered

This first week provides an introduction to the course, schedule, registration in the course in the AWS Academy on Canvas, and our first AWS Academy Course module..

AWS Academy Topics

AWS Module 1 Welcome to AWS Academy Machine Learning Foundations introduces us to the course objectives, prerequisites, and topic outline, providing a brief overview of each course module. The module then describes professional areas that require machine learning expertise and general guidelines for the types of skills that we need to develop to pursue these different professions. The module wraps up with links to resources, documentation, and white papers, We will also find a valuable resource in the Introduction section, Trends in Cloud Computing, which provides an overview of some of the latest trends in cloud computing at AWS, along with resources that we can use to stay informed.

Activities

Suggested Participation Total: 4.5 Hours

Includes meetings, learning resources in the AWS Academy, study and topic research based on suggested web resources listed in the course modules

Online Live Session: 3 hours

Course Syllabus: 0.5 hours

AWS Academy Content Time: 1 hour

• Pre-Course Survey (5 min)

• Trends in Cloud Computing (25 min)

• Module 1 Video Lessons and Student Guide (30 min)

Objectives

At the end of this module, you should be able to:

• MO1. Identify course prerequisites and objectives.

• MO2. Describe the various roles that require machine learning knowledge.

• MO3. Identify resources for further learning.

This week, we provide a broad overview of machine learning (ML) concepts, tools, and terminology. We are also introduced to the machine learning pipeline used throughout much of the course.

AWS Academy Topics

AWS Module 2 Introduction to Machine Learning begins with an overview of ML and generative AI and how it relates to the general field of artificial intelligence and deep learning. It summarizes the flow of ML, along with a timeline for its growth from traditional computing. The module then surveys business problems that can be solved using ML, how ML problems can be group into learning categories (reinforcement, supervised, unsupervised), and reviews typical applications including computer vision and natural language processing (NLP). The module then introduces the ML pipeline, a general model for the development phases of a ML application that we will apply throughout the course. Details for the various stages are presented in later modules. This is followed by descriptions of some of the more common tools used for developing ML applications, including Amazon Sagemaker, Jupyter Notebook and JupyterLab, pandas, NumPy, and scikit-learn, and a demonstration of the Amazon SageMaker console. The module wraps-up by discussing some of the major challenges for development efforts in ML.

Activities

Suggested Participation Total: 7.0 Hours

Includes meetings, learning resources in the AWS Academy, study and topic research based on suggested web resources listed in the course modules, and a short ten-question knowledge check.

Online Live Session: 3 hours

AWS Academy Content Time: 4 hours

• Module 2 Video lessons (45 min)

• Sandbox practice activity (1 hr)

• Student guide and associated weblinks (2 hours)

• Deliverables Total: 100 points

  • Knowledge Check 2 (10 min) - 100 points

Objectives

At the end of these modules, you should be able to:

CO1. Describe machine learning.

• MO1. Recognize how machine learning, deep learning, and generative artificial intelligence (AI) are part of artificial intelligence.

• MO2. Describe artificial intelligence and machine learning terminology.

• MO3. Identify how machine learning can be used to solve a business problem.

• MO4. Describe the machine learning process.

• MO5. List the tools available to data scientists.

• MO6. Identify when to use machine learning instead of traditional software development methods.

This week we begin our three-week coverage of the ML pipeline using Amazon SageMaker, with an emphasis on supervised classification problems, although the techniques can be used for other ML problem domains.

AWS Academy Topics

Module 3: Implementing a Machine Learning Pipeline with Amazon SageMaker

• Section 1: Formulating machine learning problems

• Section 2: Collecting and securing data

• Section 3: Evaluating your data

In section 1, we begin by introducing how to turn a business problem into an ML problem by providing an example based on detecting fraudulent transactions, as well as introducing the datasets used in the module. In section 2, we are introduced to collecting and securing our data, both organizational and open source. The storage and extract, transform, and load (ETL) process introduces AWS storage services that can store data for ML, however, for this course only Amazon S3 is used. The advantages of scripting solutions is discussed, followed by security considerations, and the business and ethical obligations of keeping data secure and providing audit information. From here, we have our first hands-on lab, Amazon SageMaker - Creating and importing data. In section 3, we focus on evaluating data, learning how to use pandas to investigate a dataset. In the accompanying lab, Amazon SageMaker - Exploring Data, we are introduced to the scenario that is used in most of the labs for this module, where we will load the data and use pandas to visualize statistics.

Activities

Suggested Participation Total: 5.75 Hours

Includes meetings, learning resources in the AWS Academy including hands-on labs, study and topic research based on suggested web resources listed in the course modules.

Online Live Session: 3 hours

AWS Academy Content Time: 2.75 hours

• Module 3, Sections 1 - 3 Video Lessons (30 min)

• Lab 3.2 - Amazon SageMaker - Exploring Data (30 min)

• Student guide review (30-60 min.)

• Deliverables Total: 100 points

  • Lab 3.1 - Amazon SageMaker - Creating and Importing Data (40 min) - 100 points 

Objectives

At the end of this module, you should be able to:

CO2. Implement a machine learning pipeline using Amazon SageMaker.

• MO1. Formulate a problem from a business request.

• MO2. Obtain and secure data for machine learning (ML).

• MO3. Build a Jupyter notebook by using Amazon SageMaker.

• MO4. Outline the process for evaluating data.

This is our second week of our three-week coverage of the ML pipeline using Amazon SageMaker.

AWS Academy Topics

Module 3: Implementing a Machine Learning Pipeline with Amazon SageMaker

• Section 4: Feature engineering

• Section 5: Training

• Section 6: Hosting and using the model

In section 4, we are introduced to feature engineering, first reviewing the different strategies for extraction and selection, then learning about missing data, dealing with outliers, and encoding non-ordinal data. Similar to section 3, we will take a look at some code examples. Feature selection begins the iterative processes of preparing data and training a model, and coverage of these processes will be repeated over the next few sections. Next, in our lab, Amazon SageMaker - Encoding Categorical Data, we practice encoding categorical variables, using a notebook instance provided for this purpose. The goal of section 5 is to introduces us to model training, looking at file formats, data splitting and the key topic, cross-validation. A major goal of this section is gaining an understanding of which algorithms to try for a given problem. A demonstration of splitting and training data is provided, followed by the lab, Amazon SageMaker - Training a model, which continues to explore the dataset used in the previous lab. In section 6, we cover hosting and using the model, preparing a hosting environment on Amazon SageMaker, with a discussion of the advantages of using a managed hosting environment. Additional code examples are presented as well. In our lab, Amazon SageMaker - Deploying a model, we continue our work with the same dataset, deploying a trained model created for us and performing a prediction against it.

Suggested Participation Total: 6.25 Hours

Includes meetings, learning resources in the AWS Academy including hands-on labs, study and topic research based on suggested web resources listed in the course modules.

Online Live Session: 3 hours

AWS Academy Content Time: 3.25 hours

• Module 3, Sections 4 - 6 Video Lessons (45 min)

• Student guide review (30-60 min.)

• Lab 3.4 - Amazon SageMaker - Training a model (30 min)

• Deliverables Total: 200 points

  • Lab 3.3 - Amazon SageMaker - Encoding Categorical Data (30 min) - 100 points 

  • Lab 3.5 - Amazon SageMaker - Deploying a model (30 min) - 100 points

Objectives

At the end of this module, you should be able to:

CO2. Implement a machine learning pipeline using Amazon SageMaker.

• MO5. Explain why data must be preprocessed.

• MO6. Use open source tools to examine and preprocess data.

• MO7. Use Amazon SageMaker to train and host an ML model.

In this third week of our three-week coverage, we wrap up our introduction to the ML pipeline using Amazon SageMaker.

AWS Academy Topics

Module 3: Implementing a Machine Learning Pipeline with Amazon SageMaker

• Section 7: Evaluating the accuracy of the model

• Section 8: Hyperparameter and model training

Section 7 covers evaluation of a model's performance, and owing to our focus on classification, examining a confusion matrix and the metrics that can be calculated from it. We also discuss the concept of success, what it means for a business problem, and how different objectives can change the goals of the metrics. The student guide contains a few examples about detecting cats and diagnosing heart disease patients. In the associated lab, Amazon SageMaker - Generating model performance metrics, we get a chance to evaluate the accuracy of the model we trained in the prior lab, while calculating metrics from the test data results. In section 8, we cover our last topic from this module, hyperparameter tuning, and are also provided a quick look at Amazon SageMaker Autopilot. Hyperparameter tuning is a key task for creating the best model. Now that we understand metrics, we learn how a tuning job can be automated to find the best metrics. Understanding the hyperparameters isn’t the goal here—the goal is understanding the process of tuning. To tune a model, we must be able to find the right parameters to tune in the documentation, and apply that in code. If we understand the metrics section covered earlier, how auto tuning works will make sense. The last part of this section demonstrates AutoML by using Amazon SageMaker Autopilot. The section is followed by demonstrations covering hyperparameter optimization and and how to use the SageMaker Autopilot function. The section concludes with the lab, Amazon SageMaker - Hyperparameter tuning, wrapping up the module.

Suggested Participation Total: 6 Hours

Includes meetings, learning resources in the AWS Academy including hands-on labs, study and topic research based on suggested web resources listed in the course modules.

Online Live Session: 3 hours

AWS Academy Content Time: 3 hours

• Module 3, Sections 7-8 Video Lessons (45 min)

• Student guide review (30-60 min.)

• Deliverables Total: 300 points

  • Lab 3.6 - Amazon SageMaker - Generating model performance metrics (30 min) - 100 points 

  • Lab 3.7 - Amazon SageMaker - Hyperparameter tuning (30 min) - 100 points

  • Knowledge Check 3 (10 min) - 100 points

Objectives

At the end of this module, you should be able to:

CO2. Implement a machine learning pipeline using Amazon SageMaker.

• MO8. Use cross-validation to test the performance of an ML model.

• MO9. Use a hosted model for inference.

• MO10. Create an Amazon SageMaker hyperparameter tuning job to optimize a model’s effectiveness.

AWS Academy Topics

Module 4: Introducing Forecasting

The purpose of this week's module is to introduce us to forecasting, starting with an overview before covering how to handle time series data. The module ends by using Amazon SageMaker Canvas to make a prediction. In our overview, we will look at different use cases, followed by walking us through examining and understanding the challenges of working with time series data, like missing data, and the concept of lookahead using future values to inform the past, and the negative affect that can have on a model. We also discuss timestamps, frequency consistency, and using up- and downsampling to improve the data and model, and the concept of spurious correlations. We are then walked through the use of SageMaker Canvas, a managed service, for forecasting. The module also includes a simulation in place of a lab, Creating a Forecast with Amazon SageMaker Canvas, where we assume the role of a business analyst working for a consumer electronics company, using historical time-series data to build a model

which can be used to forecast demand for electronic items.

Suggested Participation Total: 5 Hours

Includes meetings, learning resources in the AWS Academy including hands-on labs, study and topic research based on suggested web resources listed in the course modules.

Online Live Session: 3 hours

AWS Academy Content Time: 2 hours

• Module 4 Video Lessons & Demos (35 min)

• Simulation: Creating a Forecast with Amazon SageMaker Canvas (15 min)

• Student guide review (30-60 min.)

• Deliverables Total: 100 points

  • Knowledge Check 4 (10 min) - 100 points

Objectives

At the end of this module, you should be able to:

CO2. Use managed Amazon ML services for forecasting.

• MO1. Describe the business problems solved by using machine learning forecasting.

• MO2. Describe the challenges of working with time series data.

• MO3. List the steps that are required to create a forecast by using Amazon SageMaker Canvas.

• MO4. Use Amazon SageMaker Canvas to make a prediction.

AWS Academy Topics

Module 5: Introducing Computer Vision

This week's module introduces us to machine learning for computer vision and it's use cases. The module begins by describing common applications of computer vision, such as autonomous driving and medical imaging, explaining the technical challenges of building computer vision applications, and describes bounding boxes and confidence scores. We then look at three different use cases for the Amazon Rekognition service, the overall workflow, and see a demonstration of the service. We wrap up the module with Lab 5 - Guided Lab: Facial Recognition, using Amazon Rekognition to perform facial recognition of known faces in a photo.

Suggested Participation Total: 5.5 Hours

Includes meetings, learning resources in the AWS Academy including hands-on labs, study and topic research based on suggested web resources listed in the course modules.

Online Live Session: 3 hours

AWS Academy Content Time: 2.5 hours

• Module 5 Video Lessons & Demos (1 hour)

• Student guide review (30-60 min.)

• Deliverables Total: 200 points

  • Lab 5 - Guided Lab: Facial Recognition (30 min) - 100 points

  • Knowledge Check 3 (10 min) - 100 points

Objectives

At the end of this module, you should be able to:

CO4. Use managed Amazon ML services for computer vision.

• MO1. Use cases for computer vision.

• MO2. Describe the Amazon managed machine learning (ML) services for image and video analysis.

• MO3. List the steps required to prepare a custom dataset for object detection.

• MO4. Describe how Amazon SageMaker Ground Truth can be used to prepare a custom dataset.

• MO5. Use Amazon Rekognition to perform facial detection.

AWS Academy Topics

Module 6: Introducing Natural Language Processing

This week's module introduces us to natural language processing (NLP), starting with an overview that describes the types of problems that can be solved with NLP solutions, introducing terminology and concepts. NLP predates machine learning as field, so although the module uses the same ML flow to describe the steps for developing NLP solutions, some of the NLP domain terminology is distinct. Preprocessing for NLP is revisited, describing some of the specific step that must be taken during data collection and labeling, including unique requirements due to text-based data. Several NLP managed services are demonstrated, including Amazon Polly, Translate, and Comprehend. The module wraps up with Lab 6 - Amazon Lex - Create a Chatbot, where we create a simple bot that can make dental appointments.

Suggested Participation Total: 6 Hours

Includes meetings, learning resources in the AWS Academy including hands-on labs, study and topic research based on suggested web resources listed in the course modules.

Online Live Session: 3 hours

AWS Academy Content Time: 3 hours

• Module 6 Video Lessons & Demos (1 hour)

• Student guide review (30-60 min.)

• Deliverables Total: 200 points

  • Lab 6 - Amazon Lex - Create a Chatbot (1 hour) - 100 points

  • Knowledge Check 6 (10 min) - 100 points

Objectives

At the end of this module, you should be able to:

CO5. Use managed Amazon ML services for natural language processing.

• MO1. Describe the natural language processing (NLP) use cases that are solved by using managed Amazon ML services.

• MO2. Describe the managed Amazon ML services available for NLP.

• MO3. Use managed Amazon ML services.

AWS Academy Topics

Module 7: Introducing Generative AI

This week's module provides an overview of generative artificial intelligence (AI) concepts, tools, and terminology, describing and defining generative AI, while summarizing its application to a broad range of use cases, including enhancing customer experiences, boosting employee productivity, optimizing processes, and enhancing creativity and content creation. It then introduces us to foundation models (FMs), comparing them to traditional ML models, and providing a brief overview of the tasks that a foundation model can perform. It also includes an overview of large language models (LLMs) and prompt engineering. A demonstration explores FMs in SageMaker Studio, highlighting its use to create images from text prompts. The module then describes AWS generative AI offerings and their benefits. Amazon's AI coding companion, Q Developer, is introduced as a tool for speeding up development of common repetitive tasks, and is used to demonstrates creating a linear regression model with Jupyter Notebooks for data science and ML tasks.

Module 8: Course Wrap-up

Our last module summarizes our course, the objectives, and links to AWS documentation and whitepapers, and provides a roadmap for working toward certification.  

Suggested Participation Total: 6 Hours

Includes meetings, learning resources in the AWS Academy including hands-on labs, study and topic research based on suggested web resources listed in the course modules.

Online Live Session: 3 hours

AWS Academy Content Time: 3 hours

• Module 7 & 8 Video Lessons & Demos (1 hour)

• Student guide review (30-60 min.)

• Deliverables Total: 100 points

  • Knowledge Check 6 (10 min) - 100 points

Objectives

At the end of this module, you should be able to:

CO6. Identify how Amazon ML services for generative AI are used.

• MO1. Define generative AI.

• MO2. Differentiate between traditional ML and generative AI.

• MO3. Use a Foundation Model (FM) in a Jupyter notebook to generate an image from text.

• MO4. Identify the benefits of Amazon Q Developer

• MO5. Explain how Amazon Q Developer works with Amazon SageMaker to speed-up development of your ML application.

AWS Module 3 Implementing an ML Pipeline with Amazon SageMaker

Duration: 30 minutes

OVERVIEW

In this guided lab on Amazon Sagemaker, you will learn how to launch an Amazon SageMaker notebook instance. From that instance, you will learn how to create a Jupyter notebook. You will learn how to create code and Markdown cells within the notebook. You will download data from an external source, then learn how to save your notebook locally so you can continue working on labs across sessions.

Amazon SageMaker is a fully managed service that provides every developer and data scientist with the ability to build, train, and deploy machine learning (ML) models quickly. Amazon SageMaker removes the heavy lifting from each step of the machine learning (ML) process to make it easier to develop high quality models.

Objectives

• Launch an Amazon SageMaker notebook instance

• Launch a Jupyter notebook

• Run code in a notebook

• Download data from an external source

• Upload and download a Jupyter notebook to your local machine

AWS Module 3 Implementing an ML Pipeline with Amazon SageMaker

Duration: 30 minutes

OVERVIEW

In this lab, you will examine the data loaded from the vertebral column dataset that you downloaded in the previous lab. 

Objectives

• Explore and display statistics by using pandas

• Use charts to explore data characteristics

AWS Module 3 Implementing an ML Pipeline with Amazon SageMaker

Duration: 30 minutes

OVERVIEW

In this lab, you will encode categorical variables. You will use the automobile dataset from the UC Irvine Machine Learning Repository. This dataset works well because it contains many categorical features.

The lab environment contains a notebook instance, which students connect to. The instance contains the notebook that students will open to complete the lab. Again, students will review the Markdown and run the code.

Objectives

• Encode ordinal categorical data

• Encode non-ordinal categorical data

AWS Module 3 Implementing an ML Pipeline with Amazon SageMaker

Duration: 30 minutes

OVERVIEW

In this lab, you will continue exploring the biomechanical vertebral column dataset. You will first split the dataset into three separate datasets for training, validation, and testing. You will then use this data to train a machine learning (MML) model by using the XGBoost algorithm.

The lab environment contains a notebook instance, which students connect to. The instance contains the notebook that students will open to complete the lab. Again, students will review the Markdown and run the code.

Objectives

• Split data into training, validation and test datasets

• Train a XGBoost model in Amazon SageMaker

AWS Module 3 Implementing an ML Pipeline with Amazon SageMaker

Duration: 30 minutes

OVERVIEW

Hosting and consuming a model: In this lab, you will deploy a trained model and perform a prediction against the model. You will then delete the endpoint and perform a batch transform on the test dataset.

The lab environment contains a notebook instance, which students connect to. The instance contains the notebook that students will open to complete the lab. Again, students will review the Markdown and run the code.

Objectives

• Deploy a machine learning model

• Use the test dataset to perform a batch transformation with the model

AWS Module 3 Implementing an ML Pipeline with Amazon SageMaker

Duration: 30 minutes

OVERVIEW

Evaluating Model Accuracy: In this lab, you will evaluate the model that you trained in previous labs and calculate metrics based on the results of the test data.

The lab environment contains a notebook instance, which students connect to. The instance contains the notebook that students will open to complete the lab. Again, students will review the Markdown and run the code.

Objectives

• Use the test data to generate predictions

• Generate a confusion matrix from the results

• Generate performance metrics for the model

AWS Module 3 Implementing an ML Pipeline with Amazon SageMaker

Duration: 30 minutes

OVERVIEW

In this lab, you will create a hyperparameter tuning job to tune the model that you created previously. You will then compare the metrics of the two models.

The lab environment contains a notebook instance, which students connect to. The instance contains the notebook that students will open to complete the lab. Again, students will review the Markdown and run the code.

Objectives

• Use Amazon SageMaker to create a hyperparameter training job

• Tune an XGBoost model by using Amazon SageMaker

• Test the tuned model by using performance metrics

AWS Module 5 Introducing Computer Vision (CV)

Duration: 30 minutes

OVERVIEW

This lab is designed to reinforce the concept of leveraging an AWS-managed database instance for solving relational database needs.

Amazon Relational Database Service (Amazon RDS) makes it easy to set up, operate, and scale a relational database in the cloud. It provides cost-efficient and resizable capacity while managing time-consuming database administration tasks, which allows you to focus on your applications and business. Amazon RDS provides you with six familiar database engines to choose from: Amazon Aurora, Oracle, Microsoft SQL Server, PostgreSQL, MySQL and MariaDB.

Objectives

• Create a custom collection for Amazon Rekognition

• Add an image to a custom collection

• Detect known faces in an image

General Steps

• Create a collection

• Upload an image of a face that will be used for detection to the SageMaker Jupyter notebook

• Add the image to the collection

• View the bounding to be created for the image

• List the faces in the collection

• Use the collection to find a face

• View the bounding box of the face

• Delete the collection 

AWS Module 6 Introducing Natural Language Processing

Duration: 60 minutes

OVERVIEW

Natural language processing (NLP) is a common use case for machine learning (ML).

In this lab, you will learn how to use Amazon Lex V2 to create a chatbot that users can interact with to create a dental appointment.

Objectives

• Create and configure test a bot by using Amazon Lex

• Create a Lambda function and configure it to work with Amazon Lex

• Create a static webpage in Amazon S3 to host the bot

• Interact with the bot in the webpage

General Steps

• Create and configure a bot using the Amazon Lex ScheduleAppointment example

• Create an AWS Lambda function to handle dialog and fulfillment tasks.

• Configure the MakeAppointment intent to use your Lambda function for dialog management and fulfillment

• Test the bot in the console and through a static webpage hosted on Amazon Simple Storage Service (Amazon S3)