Machine Learning Projects Employers want to see

Getting into machine learning, then doing cookie-cutter projects and following basic tutorials is like trying to win a formula 1 race in a Go-Kart.

You will go, but you won't compete, and you certainly won't win.

I've reviewed hundreds of ML portfolios and interviewed many other students in real science and ML roles, and I can tell you this: People who get hired building projects pass tutorials.

Therefore, in this article, I will break down the exact types of projects and specific frameworks in global negotiations and job offers.

It's not easy.

But that's why they work.

Retrieving a research paper

Think about it.

The mechanical research paper is the culmination of several months of work by some of the leading workers in the field, condensed into a few pages of text.

The amount of information in these papers is great.

So, if you get down, take out and recycle these papers yourself, think how much you will learn.

It's like trying to rebuild a formula 1 car from Blueprints – you may not have the same tools as the original engineer, but with an understanding of each and every BOLT. And when you finally get your version working, you'll understand racing on a level that most people never reach.

Recycling paper will teach you so many skills:

• Ability to understand complex calculations associated with cutting edge models.
• Ability to use Moshiated models using code from scratch or simple libraries.
• Be able to think creatively and apply your knowledge to new ideas.

And the important part is that most, and I mean almost 99%, candidates do not do this, so you will be able to get out quickly.

However, it is not easy, and I can tell you that from first hand experience. But it simply won't hire you these days.

Now, how you go about using the paper could be a whole post in itself, but let me run you through the key steps:

1. Read the paper. After that, I ask, come back again, until you fully understand what the paper was trying to solve, the algorithm used, the details, and why the results were important and why they were heard or expected. Depending on your experience, this may take a while.
2. If you don't understand some concepts, go study. This is not a waste of time, as you actively fill in the knowledge gaps you have.
3. Sketch / code advanced design, such as inputs and outputs, complex system design and ML model design.
4. Start using the simple part and find it.
5. Build a rough working prototype.
6. Adjust and try to replicate the results.

Some papers I recommend using:

This is mostly within the deep learning space, but you can find papers that are relevant to the field you want to study.

Some useful websites for finding papers:

Solve your problem

“What projects should I build”?

This is the second most common question I get asked, the first being how I got it right!

This thing, most people do not understand that the question is whether it is wrong to ask (one project, not a good question).

If I gave you a specific project to do, there wouldn't be a story behind it in the conversation.

What will you say?

“A man from the Internet said I should be able to build”

Not exactly a great situation to be in.

The project that will come out is a secret to you personally, and you are motivated to solve it. That is very good and interesting, and it will show during the interview.

For example a project

Let me give you an example of a good project.

I mentioned this story in a previous post, but I'll repeat it to really emphasize the type of projects you should build.

At my previous meeting, we were hiring a junior Data Scientist to work for us Work Research Problems.

The candidate we ended up hiring had a project going on that was directly related to the job and was a problem they were interested in solving.

They were interested in Fantasy Football (NFL) and designed their own algorithm to better provide player picks each week.

They even go on to read journal papers on other people's solutions and use other ideas. See link for research papers!

My outline

Here is a simple outline you can follow to come up with a project similar to the one I just mentioned.

1. List at least five things you like outside of work.
2. For each topic, write down five questions that you would be interested in answering or solving. So, in total, you will have 25 possible questions.
3. Now, think about how machine learning can help answer those questions. Don't worry if the question doesn't seem difficult; name it. But of course, don't try to create robot dogs or something!
4. Finally, choose one question that interests you the most. Ideally, choose something that is just within your reach; That way, you'll really learn and push yourself out of your comfort zone.

This exercise will take you 10 minutes, so you have no reason not to, and it will give you a project idea that will help you get to work.

Problem formulation and measurement

However, the idea itself will not be enough. For that, the project needs a certain difficulty and level.

This can be shown and expressed in different ways.

• You can deploy the project end-to-end using production code, cloud systems such as AWS and contain an algorithm that uses An artist and Kubernetes.
• You can use a really complex algorithm, for the world. Reading research papers is perfect for this!
• You can make it so that users can interact with the project, like an online application.
• You can make it solve various problems, such as a suite of models.

There are so many options, and it's easy to get overwhelmed.

Start and learn as you go. That's all you need to do.

Some Ideas

If, for some reason, you don't think to do the above, even if they will, here is a list of additional project ideas.

• Ask AI about a project; Give it a quick quickie, of course.
• Enter the kaggle competition, but you need to rank well to stand out.
• Use an AI / foundational model to solve a personal problem.
• Code Machine Learning ALGORITHMS from scratch using only basic, or better, traditional Python.

Now, if you want me to do it manually, here's a list of more granular projects to try:

• Emphasis on Learning PAC-MAN or any other game.
• Building a language model from scratch.
• A computer vision model for classifying images of anything.
• Sentiment analysis on a social media platform for a specific topic.
• Recommendation program for your favorite app.
• Good LLM planning for a specific use case.

Also, I offer high-quality ideas because these needs to be personal to really stand out.

After you've created these projects, you're ready to start applying for jobs!

But for real country talks, you'll need a rock solid resume.

So what makes the difference between an overlooked and noticed resume?

Find out in my previous post below.