Zeu Park

A curious developer documenting my journey through code, from Python basics to machine learning and beyond

2 minute read

Introduction

I set out to explore an ambitious project:
“Why is machine learning so hard for beginners?”

It felt important. It was personal. It was big.
But as I tried to plan it, I realized I had made a critical mistake — I jumped into a complex, research-heavy topic without having the technical foundation, tools, or time to handle it. Here’s an honest reflection on why I’m putting it on pause and what I’ve learned.

Domain Limitations

The core issue was that I chose a domain I couldn’t yet understand deeply.

I was asking, “Why is ML hard?” before I had truly experienced machine learning myself.

I lacked hands-on experience with:

  • Building even a basic model
  • Understanding real learning pain points in context
  • Working with real ML datasets

Even industry professionals tackle these questions after years of practice, and they usually rely on structured collaboration with experts and data engineers. I didn’t yet have those support systems — and it showed.

Technical Limitations

There were too many things I didn’t yet know:

  • How to crawl or collect meaningful data
  • How to clean, reshape, or verify it using tools like pandas or NumPy
  • How to decide if a dataset is even usable or relevant

I’m still learning the basics of:

  • Python data handling
  • Exploratory data analysis (EDA)
  • Data transformation techniques
  • Evaluation of dataset quality

I simply wasn’t ready to make technical decisions about a domain-heavy project.

Time Limitations

This project wasn’t my full-time job — I’m learning Python, SQL, and ML as part of ongoing classes.
Trying to juggle all that while designing a research-grade project in my free time was unrealistic.

Even just trying to understand the domain meant reading books like:

  • Python for Machine Learning & Pandas
  • Web Scraping with Python
  • Dacon Competition Winning Solutions
  • Introduction to Machine Learning with Python

But reading is slow, and deep research requires focus. I didn’t have enough time to do it all well.

Reflection

This was a humbling experience.

I tried to do something beyond my current ability — and that’s okay.

The project was:

  • Too large in scope
  • Too technical for my current level
  • Too time-intensive to balance with other commitments

But I learned more about project selection, self-assessment, and setting achievable goals than I ever could have by pretending it was going fine.

What I’ll Do Next

  • Refocus on Kaggle community competitions to observe and learn real workflows
  • Pick topics I’m already familiar with to narrow the domain
  • Start small, with structured problems I can realistically solve
  • Write new project proposals based on my actual skills
  • Continue reading and solving problems to widen my technical comfort zone

Final Thought

Ambition is good. But so is realism.

This project was never a failure — it was a checkpoint.
And now, I’m ready to move forward more grounded, more honest, and more prepared.

Tags:

Updated:

You May Also Enjoy