An Uncommon But Efficient Take on Passing the AWS Certified Machine Learning Specialty Exam

A guide for experienced ML practitioners who don’t feel like studying AWS services for months only to acquire that pretty badge.

Photo by Rubén Bagüés on Unsplash

Having worked with AWS services for eighteen months, I determined it was time to pursue the AWS Certified Machine Learning Specialty certification. However, I hesitated to proceed for months, as many guides suggested taking other entry-level AWS exams first or following lengthy preparation schedules, which I deemed a waste of time and money.

It was then that I concluded that my experience in delivering practical ML projects had equipped me sufficiently, and I devised a study plan to quickly and effectively pass my first professional certification exam.

AWS Certified Machine Learning Specialty Badge. Image by AWS.

I will share my approach and other important insights about the exam in the following text. Note that my perspective is based on my personal experience from March 2023.

Content and question style

The exam guide provided by the official source covers 95% of the content that will be tested, and you can expect to encounter a little more. The passing score for the exam is 750, with a maximum possible score of 1000.

The exam will consist of 65 questions, covering Data Engineering (20%), Exploratory Data Analysis (24%), Modeling (36%), and Machine Learning Implementation and Operations (20%). The questions will be in multiple-choice (with one correct option) and multiple-response (with two or more correct options) formats.

To succeed in the exam:

• Focus on selecting the best option since there may not be an entirely correct answer. For example, suppose you must choose the optimized metric in a hyperparameter tuning job of a cancer classifier. In that case, you might have to go with F1-Score even though recall is the preferred metric since false negatives are highly costly.
• Reading the questions carefully is crucial, as every word matters. In complex architectural scenarios, various choices can correctly solve the problem, but you must pick the one that better meets the requirements.
• A high-level knowledge of AWS services is essential for success, but you do not need years of experience. It is sufficient to know the value offering, why it is valuable, and the best use cases for each. Hence, there is no need to read through the API docs.
• There are also more non-AWS-specific conceptual machine learning questions than you might think. This can be viewed as a blessing!

Study plan

Firstly, to be adequately prepared, you should be confident in your understanding of ML modeling and problem-solving and have a grasp of model operationalization and data preparation. Though hands-on experience is not necessary for the AWS portion, it’s ideal to have familiarity with their core offerings to simplify the process.

Following this strategy, I achieved a solid 880 score on the MLS-C01 exam. Image by author.

Here is a breakdown of my preparation plan:

1. Begin by reading the official exam guide and attempting the official sample questions to familiarize yourself with the exam style. Don’t be discouraged if you don’t perform well initially, as I only scored 5/10 on my first attempt.
2. Next, enroll in the AWS Certified Machine Learning Specialty 2023 — Hands On! course on Udemy, which is a comprehensive 12-hour course. You can skip parts you’re already comfortable with, but ensure you can complete the quizzes at the end of each chapter and score at least 8/10 on the Warmup Test.
3. If you’re having trouble with the course questions, dedicate some time to studying the concepts you’re getting wrong until you feel confident you can answer related questions correctly.
4. Take the AWS Certified Machine Learning Specialty Full Practice Exam on Udemy, ensuring you complete it within 2 hours with at least 80% accuracy.
5. Proceed to the AWS Certified Machine Learning — Specialty Official Practice Question Set, which provides the closest approximation to the actual exam. In my experience, this was even more challenging, so aim to score at least 70%.
6. Additionally, attempt the AWS Exam Readiness — Module 7: Study Questions, which I found to be equally challenging.
7. Finally, review your practice exams a day or two before the exam, noting the topics you feel less confident about. Focus your efforts on those areas to ensure your preparation is complete.

It’s worth noting that while some resources require payment, the cost is quite affordable — I spent around US$10.00 in total.

Conclusion

While the AWS Machine Learning Specialty exam is challenging, it’s not an impossible task. With a decent amount of experience in Data Science, a bit of willpower, and discipline, you too can earn your badge. Let’s stop overcomplicating things and approach the exam with confidence.

I hope this article has been beneficial in assisting you in achieving your objectives. Best of luck on the exam!