My motivations and experiences for 2020 AWS Machine Learning certification, updated for 2023
Why did I start a Machine learning education?
The AWS Certified Machine Learning Specialty isn’t my first experience with machine learning. To find out why I thought it would be fun to put myself through these tough exams, check out my previous articles on machine learning and deep learning. This article is about the AWS machine learning speciality, my study path, and the AWS machine learning specialty exam. This article includes updates for 2023 on the importance of a growth mindset in the AWS exam preparation and Common Challenges and Solutions.
The AWS Machine Learning Speciality was one of the most challenging exams I’ve taken but also the most rewarding. It’s tough, so it’s a fantastic feeling when you pass. Reflecting on my achievements in light of recent technological advancements, such as ChatGPT and DALL-E 2, I am proud of my accomplishment. With cutting-edge generative AI innovations coming out at a rate that seems impossible to stop, we are entering a new era of creativity and business innovation.
Why did I take the AWS Certified Machine Learning Specialty exam?
There are many reasons why one might choose to pursue the AWS Certified Machine Learning Specialty exam. It was a way to deepen my understanding of the rapidly evolving field of machine learning, challenge myself and achieve a personal milestone, and stay updated with the latest technologies and techniques. Overall, the exam was a valuable experience that allowed me to further my personal and professional development and stay at the forefront of this exciting field.
What is an effective strategy for preparing for the exam?
A challenging exam like the AWS machine learning speciality requires technical and mental preparation. Here are some tips to help you approach your exam preparation with confidence and resilience.
Embrace challenges: Don't shy away from difficult topics. Challenges provide opportunities to learn and grow. Instead of being discouraged by failure, use it as an opportunity to learn from your mistakes and improve your approach.
Be open to feedback: Feedback is a valuable tool for learning and growth. Be open to receiving feedback from others, including peers and mentors. Use feedback to identify areas of improvement and adjust your approach.
Stay motivated: Setting achievable goals, breaking down your study plan into manageable chunks, and tracking your progress can help you stay motivated. I used Trello boards and whiteboards to track my progress for the exam domains and work through challenging sections of the courses.
Practice, Practice, Practice: Practice makes perfect, and the more you practice, the more comfortable and confident you will feel on exam day. Use online resources, practice exams, and hands-on experiences to reinforce your knowledge and skills. I took the practice exams multiple times until I felt confident I had learned the topics. Additionally, I challenged myself to build a machine-learning pipeline from memory to practice.
Everyone's learning journey is unique; what works for one person may not work for another. So be bold and try different approaches and methods until you find the best.
Why did I choose the AWS Platform?
I chose AWS because I was already familiar with it. I’d used AWS in large corporates to migrate workloads into the cloud and in a start-up role to build a cloud-native bank from the ground up. I’ve used products like SageMaker in my machine learning education with courses from FastAI, Stanford and Coursera. However, it seemed fitting AWS should be my first platform-specific certification.
What topics do you need to learn?
In Amazon’s words, the AWS Certified Machine Learning Speciality Certification “validates a candidate’s ability to design, implement, deploy, and maintain machine learning (ML) solutions for given business problems.” To cover all that, the exam is split into four domains. The percentage is how much of the exam is made up of that domain.
Domain 1 is data engineering 20%: This covers the AWS big data stack; including Streaming Data Tools (Glue & Kinesis), Storage (S3 & RDS) and analytics (Kinesis & Athena). This domain shows how these components fit within the Sagemaker ecosystem.
Domain 2 is exploratory data analysis 24%: This covers how you prepare data for modelling, feature engineering and data analysis, how you handle missing values, unbalanced data and normalisation. This isn’t AWS specific and will require a broader knowledge of exploratory data analysis.
Domain 3 is modelling 36%: This is the most significant part of the exam and requires broad knowledge of SageMaker Machine Learning Models and AWS Machine Learning Products. This domain covers choosing a suitable model for a business problem, training models, hyper-parameter tuning, model evolution, to name a few.
Domain 4 is machine learning Implementation and Operations 20%: This covers security, deployment and optimisation of solutions using SageMaker.
How Did I Approach Studying for the Exam?
As I had previous experience in machine learning, my preference was to find courses that helped me focus on the areas on AWS I needed to brush up on and any gaps that needed to be filled. instead of learning from scratch. The best method I discovered was Frank Kane’s AWS Certified Machine Learning Specialty. Frank has updated the course for 2023 with the latest SageMaker features and new AWS ML Services. I’ve taken Frank’s classes before and I like his down to earth style. He has a career history in AWS and IMBD and easily relates machine learning concepts to real-world examples. If you’re coming to this without any machine learning knowledge, it will be a challenge. A point Frank confirms in his lectures.
Similar to the actual exam domains, there are four sections to Frank’s course. There are also several labs where you can try out the tools for yourself. Unlike other learning platforms, you will require an AWS account. The AWS free tier will help minimise costs but remember to shut down and delete any builds you start. If you create a machine learning training environment with a GPU, this can be very expensive if you forget to shut the environment down.
Frank also has a separate practice exam updated in 2022. This helps prove your knowledge and get a feel for the formatting for the questions. Besides Frank’s practice exams, there are two practice exams on Udemy by Abhishek Singh, updated in 2022 that are worth taking. These were handy tools to pinpoint gaps in my knowledge and highlight areas that needed more study. Practice exams are a great resource, whatever kind of study you are undertaking.
In addition to the courses, to gain a deeper understanding of the products, I would recommend reading the AWS documentation, watching AWS Videos and creating your own environments on AWS itself. There is plenty of content covering SageMaker, machine learning pipelines, data analysis amongst others.
What Are the Common Challenges and Solutions?
One of the best ways to overcome technical difficulties when studying for the AWS Certified Machine Learning Speciality exam is to start with introductory courses or tutorials. In my previous articles on machine learning and deep learning, I have shared some great resources for beginners looking to build a strong foundation in these concepts.
Another challenge when studying for the exam is the cost of resources, mainly when using AWS Machine Learning products. To help minimise these costs, it's a good idea to plan out your study approach carefully and use the AWS free tier effectively. Additionally, you can consider using alternative, open-source tools for specific tasks. For example, I created a Lambda job to take down AWS Sagemaker instances every few hours to avoid leaving them running for extended periods.
By being strategic and thoughtful in your approach to studying for the AWS Certified Machine Learning Speciality exam, you can overcome these challenges and succeed in this exciting and rapidly growing field.
How is the AWS Exam Formatted?
The exam itself is 3 hours long and comprises 65 questions. Many are scenario-based. For example. You work for a healthcare firm and need to tune a machine learning model. What parameters will help you do this? Another example is you are working on a fraud system in financial services. How will you evaluate the effectiveness of a model? A final example is what combination of AWS Kinesis products will allow you to stream data and create visualisations.
As with all AWS exams, there are always two answers which seem appropriate, make sure you read them closely. It took me the entire 3 hours to complete the exam. Exam technique 101, always flag the questions you’re not sure of so you can return to these easily and give yourself time to check them at the end. Make use of all the time you have. There are no direct coding questions, but you may have to select when to use the correct library. For example, Scikit-learn, MLlib or TensorFlow.
What Was my Experience With the Home Proctored Exam?
Due to lockdown in the UK, the exam was home proctored. The software monitors your computer’s desktop, webcam video and audio. I’m not sure if a person sits through the entire exam, but someone did pop up on the chat. More on that later. During the exam, you can’t have any food, drink, calculator, paper or pens. There is a whiteboard in the exam software and my initial concerns about not having a calculator were unfounded.
As it was my first home proctored exam, there was a learning curve and it wasn’t without problems. Run a test a couple of days before to check the camera, mic, and operating system capability. My first challenge was the fixed iMac webcam. It wasn’t possible to move it around for the workspace check, so I bought a cheap separate webcam. The second challenge was Wi-Fi. I have good Wi-Fi, but it seemed to fail during one attempt. The advice is to connect your computer to your broadband. Once I resolved the technical issues, the proctored system worked well and I much prefer this way of taking the exams instead of travelling to a testing centre.
Once the exam is completed, you’ll know if you’ve passed or failed within a few minutes. Note the pass or fail screen is not the end of the exam!! Remember to close the last page before celebrating, assuming you pass. I did not close the last page before celebrating and this was a when rather anxious proctor popped up on the chat room instructing me to do so.
What is Explainable AI?
Explainable AI (XAI) is an emerging trend in Machine Learning that focuses on making machine learning models more transparent and interpretable. By providing insights into how AI models work and the reasons behind their output, XAI is needed to provide transparency and interpretability to machine learning models, ensuring that they produce ethical and unbiased results. AWS offers several XAI tools, such as Amazon SageMaker Clarify, to help users identify and mitigate bias and other ethical concerns in Machine Learning models. In addition, by understanding how AI models make decisions and the factors that contribute to those decisions, users can make more informed decisions about the deployment and use of their AI models.
What is Human centered AI and Design Thinking?
Human-centred AI and design thinking ensure that AI solutions prioritise customer needs and experiences. By rapidly prototyping and iterating based on user feedback, design thinking enables the development of transparent, fair, and accountable AI systems that enhance people's lives. Integrating design thinking into machine learning facilitates innovative human-centric models through practices like problem framing, empathy, ideation, prototyping, testing, implementation, monitoring, explainability, building trust, and ethics. This approach creates AI that serves humanity.
AI Ethics
For a more in-depth resources on the ethical considerations and challenges of implementing machine learning in business settings, please refer to my previous article Starting out in Deep Learning.
AWS Certified Machine Learning – Specialty
Summary
Looking back on my journey to obtaining the AWS Certified Machine Learning Certification, I can confidently say it was a challenging yet rewarding experience. Through the process, I gained an in-depth understanding of AWS products and services and expanded my machine learning knowledge while improving my python coding skills. The three-hour exam was intense, but it was an excellent opportunity to put all of my skills to the test.
In the two years since passing the exam, I have continued to apply my knowledge in various settings, which has proved incredibly valuable in my professional development. I am grateful for the experience, and I have since pursued other certifications that have expanded my knowledge and skill set even further.
Pursuing the AWS Machine Learning Specialty certification was a significant investment in my personal and professional growth. I highly recommend it to anyone interested in advancing their knowledge and skills in machine learning.
Now your on your way to passing the exam, checkout these blog posts on coding with Generative AI.
Supercharge your business with the power of machine learning. Discover how machine learning can drive innovation, enhance efficiency, and unlock new growth opportunities.
Unleash Innovation and Creativity. From transformative leadership to a growth mindset, let's apply human-centred strategies with the power of artificial intelligence in business. Together, we'll craft your business transformation. Book your 1:1 session today.
Looking for more insightful content?
Check out these related blog posts on business innovation, design thinking, AI and technology written by David Kolb.
From Innovation to Advantage: Harnessing the Power of Generative AI
Implementing Problem Solving: Strategic Approaches and Techniques for Organisational Growth
Merging Human Centered Design with Business Strategy (Video)
Enhancing the Machine Learning Pipeline with Design Thinking
Ethical Considerations for Boards in the Age of AI: Merging Technology and Values
Demystifying Artificial Intelligence for Boards: Uncovering the Power and Practicality
Navigating AI Adoption: The Essential Perspectives for Board Members
Or check out our weekly concise and valuable quick tips for visionary leaders and entrepreneurs.
Comments