Name: Towards AI Legal Name: Towards AI, Inc. Description: Towards AI is the world's leading artificial intelligence (AI) and technology publication. Read by thought-leaders and decision-makers around the world. Phone Number: +1-650-246-9381 Email: [email protected]
228 Park Avenue South New York, NY 10003 United States
Website: Publisher: https://towardsai.net/#publisher Diversity Policy: https://towardsai.net/about Ethics Policy: https://towardsai.net/about Masthead: https://towardsai.net/about
Name: Towards AI Legal Name: Towards AI, Inc. Description: Towards AI is the world's leading artificial intelligence (AI) and technology publication. Founders: Roberto Iriondo, , Job Title: Co-founder and Advisor Works for: Towards AI, Inc. Follow Roberto: X, LinkedIn, GitHub, Google Scholar, Towards AI Profile, Medium, ML@CMU, FreeCodeCamp, Crunchbase, Bloomberg, Roberto Iriondo, Generative AI Lab, Generative AI Lab Denis Piffaretti, Job Title: Co-founder Works for: Towards AI, Inc. Louie Peters, Job Title: Co-founder Works for: Towards AI, Inc. Louis-François Bouchard, Job Title: Co-founder Works for: Towards AI, Inc. Cover:
Towards AI Cover
Logo:
Towards AI Logo
Areas Served: Worldwide Alternate Name: Towards AI, Inc. Alternate Name: Towards AI Co. Alternate Name: towards ai Alternate Name: towardsai Alternate Name: towards.ai Alternate Name: tai Alternate Name: toward ai Alternate Name: toward.ai Alternate Name: Towards AI, Inc. Alternate Name: towardsai.net Alternate Name: pub.towardsai.net
5 stars – based on 497 reviews

Frequently Used, Contextual References

TODO: Remember to copy unique IDs whenever it needs used. i.e., URL: 304b2e42315e

Resources

Take our 85+ lesson From Beginner to Advanced LLM Developer Certification: From choosing a project to deploying a working product this is the most comprehensive and practical LLM course out there!

Publication

How I cleared AWS Machine Learning Specialty With Three Weeks of Preparation (I will burst some myths of the online exam)
Latest   Machine Learning

How I cleared AWS Machine Learning Specialty With Three Weeks of Preparation (I will burst some myths of the online exam)

Last Updated on July 17, 2023 by Editorial Team

Author(s): Himanshu Joshi

Originally published on Towards AI.

How I prepared for the test, my emotional journey during preparation, and my actual exam experience

Certified AWS ML Specialty Badge source

Introduction:-

I recently gave and cleared AWS ML certification on 29th Dec 2022.

Let me start by explaining why this certification is a great value add to you.

AWS ML Specialty certification tests the whole life cycle of a Data Science project. As an interviewer, it is safe to assume that one who has this certificate can easily Architect and run a Machine Learning project on his own in AWS. One cannot pass this certification if their Machine Learning basics are not strong.

So this will surely make you stand out from the crowd during interview selection.

Topics covered in this post –

  1. My background and experience working in AWS
  2. How I prepared for the certification
  3. Giving practice exams
  4. Analyzing practice exam results and covering bases accordingly
  5. Last day revision
  6. Exam day experience β€” Busting myths about the online exam (Pearson vue)
  7. Tips from my experience

My background and experience:-

I am a Lead Data Scientist at a Service based IT company with 11+ years of professional experience. I am a self-taught Data Scientist.

On the machine learning concepts side, I was very well placed.

Experience with AWS:-

I have close to a year of experience working in AWS cumulatively. As a Data Scientist, I have worked with the following services β€” S3, AWS Sagemaker, and Redshift.

2 years back I had prepared for the AWS Cloud Practioner exam, but never gave the exam, as I was not sure how much value will that certification add to my existing experience

So as you can see I was nowhere near the recommended experience given by AWS –

β€œThe target candidate is expected to have 2 or more years of hands-on experience developing, architecting, and running ML or deep learning workloads in the AWS Cloud.”

So if I can do it you can do it too with a bit of smart work ( provided you have some experience working in Data Science/ ML concepts are strong )

Preparation for Certification:-

I have a habit of over-preparing and over analyzing. Hence this time I decided to first book the exam and then start working towards it.

This I would say was the most important factor in my appearing for (and clearing) the exam, might seem a bit weird but it’s true. The human mind ( at least an engineer’s mind especially if you are from India) is trained in that way. We don’t take things seriously unless we have a deadline to meet.

So I forced myself to study by setting, a deadline.

As a complete disclosure, my company is an AWS Partner and I got a voucher from my company around 10th Dec. So I set the date for 22nd Dec and started preparing. My client being a US-based firm, and most of its employees were on leave during year-end. So this felt like the best time to prepare and take the exam.

I took AWS Certified Machine Learning Specialty 2023 β€” Hands On! by Frank Kane and Stephane Maarek β€” This was my go-to course during preparation.

This course gives a great intro on what to expect and then goes on to explain each and every aspect (depth is missing in some cases by its a great way to get the ball rolling)

Here are the components on which you are tested for the certification:-

Domain 1: Data Engineering -20%

Domain 2: Exploratory Data Analysis -24%

Domain 3: Modelling -36%

Domain 4: Machine Learning Implementation and Operations -20%

Sharing the Machine-Learning-Specialty_Exam-Guide link for reference. I would suggest going through this link and plan even before you start your preparation.

I went through these video courses as well:-

AWS Certified Machine Learning Specialty (MLS-C01) β€” for domains I was not very confident of.

Amazon SageMaker Technical Deep Dive Series β€” AWS folks have created this playlist. This was very helpful to understand the overall architecture and AWS folks thought process.

Key Takeaways:-

  1. Set a deadline and then work towards it ( Tip:- You can reschedule your exam 2 times )
  2. Just because you don’t have enough AWS experience doesn’t mean you won’t be able to clear the exam

Practice Exams:-

I gave the following practice exams:-

AWS Certified Machine Learning Specialty: 3 PRACTICE EXAMS

AWS Certified Machine Learning Specialty Full Practice Exam

AWS Certified Machine Learning Specialty (MLS-C01) -This course has 1 practice exam.

AWS Certified Machine Learning Specialty β€” Whiz Labs

These should be more than enough for you to be ready for the certification.

Analyzing practice exam results and covering bases accordingly:-

As I had mentioned earlier, I had set up my exam for 22nd Dec.

Even before I started going through the courses, I gave the 20 practice questions given by AWS. I felt very confident after this.

Post this I went through AWS Certified Machine Learning Specialty 2023 β€” Hands On!

Then I gave 2 exams just 2 days before my exam day.

I scored 63% and 72% respectively (Now I understood that I was being overconfident and needed more preparation) Many answers I got right were flukes and the questions I was confident I knew the answers to were wrong.

Tip:- The answer options given are mostly very close to each other and a word added or subtracted changes the answer drastically. So read the question and answers very very carefully.

Hence I postponed my exam by another week, but I decided not to postpone it further. ( You can postpone the exam twice, but you need to do it 24 hrs prior to the scheduled time)

Now the real preparation started for me, I researched and found even more practice tests to take and then started to diligently go through the practice test answers and explanations. I went through the AWS Certified Machine Learning Specialty 2023 β€” Hands On! again in depth. Earlier I had skimmed through it.

Analyzing the test answers and explanations teaches you way more, as only in exam situations do we realize what we know actually.

Tip:- Even if you think you don’t have any clue about the answer, use the elimination technique to discard options. In my experience, 2 options can be eliminated easily. That leaves you with a 50% chance of getting the answer right.

Note:- All practice exams were very close to the real exam apart from Whiz Labs. Whiz Labs had questions that were very lengthy and very tough. They were more from a Data Engineering angle rather than ML.

Please go through Exam Readiness: AWS Certified Machine Learning β€” Specialty β€” I had AWS Partner access (Not sure how to access without Partner access)

This acted as a very good way to revise and give practice tests, this is provided by AWS and hence the best resource.

In all the exams that I gave (around 6–8 odd practice exams), never did I comprehensively cross the 75% mark. I failed 2–3 practice tests.

I scored 66% odd in the official AWS practice test. I was dejected, to be frank, but I saw the median was around 58–60%

After all this, I was feeling very underconfident and had almost given up. But later I thought to myself, what is the worst that could happen?

I will fail β€” That’s all. So I decided to not focus on the marks and focus on learning.

This was my chance to learn AWS Architecture in depth. As the exam tests us on Data Engineering, AWS components and expects us to design solutions.

This shift in mindset did wonders. I was no longer concerned about the result but was interested to learn. I started to enjoy the process unlike earlier.

Last day revision:-

I decided to note down the observations I had from practice tests:-

Every exam I gave had new questions that I had not heard of.

Ex: Inference pipeline can deploy how many containers

I had absolutely no clue. Let’s be realistic, as a Data Scientist we hardly do deployment. We are mostly not allowed to train models on the cloud as it is costly. Given all these constraints it’s very difficult for us to master all 4 domains.

Answer to the above question β€œAn inference pipeline is a Amazon SageMaker model that is composed of a linear sequence of two to fifteen containers that process requests for inferences on data.”

The practice exams had made it clear to me that it’s not at all an easy certification. All the questions were a bit tricky.

If one doesn’t read the question very carefully anyone can make a mistake no matter what level of expertise they may have.

The exam is a 3-hour affair having 65 questions. That itself is very difficult. The exam plays on you mentally. Sitting and thinking for 3 hours is not something that we are used to nowadays.

In the practice exams, I noticed they kept all the tough questions at the start, which itself demoralises you. So even if you know the answers at the end it feels like what’s the point of answering now as I have already answered so many questions wrong.

Also as the exam progresses you are stretched so much mentally that you don’t feel like reading the questions properly. During practice tests, I felt like I should just end the exam and take a break.

So keeping all the above points in consideration, I decided to do the following on the last day:-

Play to my strengths β€”

Focus on the things I knew very well β€” EDA, ML concepts, and Modelling were my strong points. So I revised these very well.

Data Engineering and Machine Learning Implementation and Operations in AWS were my weak points. So I went through these sections in Exam Readiness: AWS Certified Machine Learning β€” Specialty and answers to the practice tests

Ignore new concepts and focus on the concepts I had learned from courses and practice tests. No point in learning new stuff when you keep forgetting already learned stuff.

Most of my underconfidence came from Whiz Labs practice exams. I was able to get 60% max in these exams.

Try to get 6–8 hrs of sleep so that I am fresh the next day.

Test the system and keep things ready as instructed by AWS.

I was able to accomplish most of the tasks apart from the sleep part (I was able to sleep hardly for 4 hours )

Exam day experience β€” Busting myths about the online exam (Pearson vue)

I had read somewhere from 1st Jan 2023 you can only take online tests. If that’s the case the next section is very important to all. As there are many myths about the online test.

While researching online, I read many myths about the exam. Pearson vue proctors are very rude, they terminate exams even if you do a small mistake, stand up, look here and there etc…

Let me tell you that most of these are only myths. The proctors are very cordial and not at all rude.

The process :-

Login 30 mins prior to the test. Do all the system tests and share pics of your surroundings.

Then you are added to a queue. There were 66 people ahead of me and it took 35–40 mins for me to start the test.

Before the exam started I spoke with the proctor and asked if I can take a washroom break. She allowed it.

About 2 hours into my exam, my network stopped working. I thought I am screwed and I will fail. But the proctor asked me to restart the system and start again. When I joined back, I apologised for the network failure. She was very understanding and asked me not to apologise as this is a technical glitch. She asked me to calm down and give the exam from where I left.

After sometime the proctor changed. The new proctor messaged me to change my camera angle so that my shoulders were also visible else he will terminate the exam. I adjusted my camera and asked him if it was fine.

This was the only instance when I was warned.

So let me assure you if you are not trying to cheat in anyway, the proctors are very cordial and not at all rude.

Results :

4–5 hours later I logged into the AWS portal and got the result.

Tips from my experience :-

  1. Set up a deadline and work towards it
  2. Take as many practice tests as possible.
  3. Don’t get disheartened by the practice test results. Going through the answers and analysing the results is very important to understand strength and weakness (AWS ML Specialty exam results are weighted score I think)
  4. Stick to your strengths
  5. Practice elimination where you don’t know the answers. This will surely increase your probability of answering correctly.
  6. 3 hours is a really long time prepare yourself mentally for it and don’t give up till the very end
  7. Don’t drink too much water on the day of exam. No toilet breaks are allowed during the exam
  8. Don’t panic come what may. The exam tests you on technical and behavioural aspects
  9. Back yourself on the exam day. You have got this. You have prepared for it and you will pass.
  10. Certification is temporary. Knowledge is permanent

All the very best. I have tried to cover my experience in as much depth as possible.

Please comment and share your experiences.

Mlearning.ai Submission Suggestions

How to become a writer on Mlearning.ai

medium.com

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming aΒ sponsor.

Published via Towards AI

Feedback ↓