⚰️

Everywhere you look, people are saying that if you're not training LLMs with billions of parameters, you're not in the profession, and linear regressions should be sent to a museum.

But now let's open analytics from airesearchtrends on the tools actually used in modern ML pipelines:
🟢 Classical ML — 47%
🟢 Neural Networks — 36%
🟢 Ensemble Methods — 26%
🟢 Clustering — 4%

Almost half of real research and production is built on classical ML. Because where interpretability, predictability, and reliability are needed, bloated neural networks lose to a well-tuned model. But the mass market has somehow decided that if you've learned to call the OpenAI API, you no longer need to understand the math under the hood. Spoiler: you can't.

In fact, that's why the next season of the Master Group is dedicated to the fundamental foundation. We move from data analysis to building models.

🗓 Starting April 25. Broadcasts will be held on Saturdays at 11:00 Moscow time.

What's on the menu:
🟢 5 broadcasts: from the very basics of machine learning to building and validating basic linear models.
🟢 1 broadcast — team practice. Stop just watching lectures. You'll work in teams on a practical task, and then we'll all review the results together (and see who messed up where).
🟢 Recordings and my code files, of course, remain with you. Take them and use them as a cheat sheet at work.

Since Telegram has been doing poorly lately, unfortunately, the closed group will be on VK.

🐍 Important filter:
This is not a "from zero to senior in two days" course. You should already be able to write Python adequately, manipulate dataframes in pandas, and handle basic data analysis (everything we suffered through in the first two seasons).

👉 If you want to jump on this train and learn the conditions, write to me in private: @obulygin91