Getting started with machine learning in 2025

🤖 Artificial Intelligence 261 views 3 replies 📌 Pinned
For those new to ML, what's the best path to take in 2025?

Should I start with:
- Traditional ML algorithms (sklearn)
- Deep learning frameworks (TensorFlow, PyTorch)
- Or jump straight into LLMs and transformers?

Looking for practical advice!

Replies (3)

Start with traditional ML algorithms first! Understanding the fundamentals is crucial:

1. Learn linear regression, decision trees, random forests
2. Then move to neural networks basics
3. Finally explore deep learning and transformers

Rushing into LLMs without foundation knowledge will hurt you later.
I recommend the fast.ai course. It takes a top-down approach - you start building models immediately, then learn the theory.

Great for staying motivated while learning!
Don't forget the math! Make sure you understand:
- Linear algebra
- Calculus basics
- Statistics and probability

You don't need a PhD, but these foundations help A LOT.
Please login or register to post a reply.