Thread by @paulabartabajo_

Wanna become a professional data scientist? 👩🏾‍🔬👨‍🔬

One that feels
- knowledgeable 🧠
- confident 😎
- and ready to charge well what she knows? 💰

Here is what you should do (spoiler alert, it is hard, but worth it) ↓↓↓

The internet is flooded with Data Science/ML content:

→ blog posts
→ newsletters
→ Twitter threads
→ Arxiv papers
→ ...

And the thing is, reading all that is not gonna get you a job.

You need to get your hands dirty ⛏️👷🏾‍♂️👷🏻‍♀️

Real learning in data science (like in life) happens when you

→ face a specific problem
→ struggle to solve it, and
→ eventually solve it.

I call this the "problem-struggle-solution cycle".

This is how you learn everything in life.
And data science is not an exception.

When you skip this cycle you feel impostor syndrome.

That is, you feel deep inside that you are not ready to become a

→ professional
→ full-time
→ and well paid

data scientist.

Impostor syndrome implies negative self-talk. And no amount of reading can cure you of this.

Is there another way to learn? 🤔

Yes, there is.

I call it "Pick-a-real-world-problem-you-care-about-build-a-data-science-solution-and-explain-it-to-others".

Also known as "Project-based learning"

For #example: let's say you wanna become an NLP engineer.

Stop reading blog posts about multi-billion-parameter Language Models.

Instead, use NLP techniques (embeddings and similarities) to build this "app that recommends what ML paper to read"↓

Starting from the end goal puts your mind in the "problem-solving" mode.

So you start asking yourself the right questions...

Question → is there any public dataset I can use?

And you find this one
kaggle.com/datasets/spsay…

Question → How do I transform the raw text into numerical features?

And you discover different "embedding" methods:

1 → TF-IDF vectorizer from scikit-learn (classic 🏛️)
2 → GloVe word embeddings (modern)
3 → Sentence Transformers using HuggingFace's library (SOTA 😎)

Question → How do I match the user query with the paper to recommend?

And you discover the cosine-similarity distance, something you probably learned in high school and suddenly re-discover 😉

Question → How do I make my app public?

And you finally build your first app with Streamlit.

Boom.

When you learn by building projects, you get 2 things:

1 → Become a problem-solver and cure your impostor syndrome.

2 → Build a portfolio. Every project you make is another valuable asset you will have when looking for jobs.

Building a real-world project is hard, but the effort is worth it.

This is why I am preparing a hands-on tutorial to help you build your OWN Machine Learning project.

Join my e-mail list to be notified when the tutorial series is out ↓
datamachines.xyz/subscribe/

Every week I share real-world Data Science/Machine Learning content.

Follow me @paulabartabajo_ so you do not miss what's coming next.

Wanna help?
Like/Retweet the first tweet below to spread the wisdom ↓↓↓

https://twitter.com/paulabartabajo_/status/1598316076344877056