Machine Learning & Data Science

What you'll learn

• You analyze data, build a model, and evaluate it honestly.
• You know where classical ML vs neural nets fit.
• You build an NLP/LLM project with fine-tuning and RAG.
• You deploy a model as an API.
• You build a GitHub/HuggingFace portfolio.

Program by modules

1. 01

   Python & data

   • numpy, pandas
   • EDA — exploratory analysis

   Project: Run EDA on a real dataset and write a report.

2. 02

   Classical ML

   • Regression, trees, gradient boosting
   • Validation without data leaks

   Project: Train and validate a model on tabular data.

3. 03

   Metrics & experiments

   • ROC-AUC vs PR-AUC, calibration
   • A/B testing basics

   Project: Compare two models by the right metric.

4. 04

   Deep learning basics

   • PyTorch, tensors, autograd
   • The training loop

   Project: Train a small neural network from scratch.

5. 05

   NLP & LLM

   • Transformers, embeddings
   • Fine-tuning (LoRA), RAG

   Project: Build a RAG system over your own documents.

6. 06

   Production

   • Inference, quantization
   • Deploy with FastAPI / vLLM

   Project: Ship a model as an API and load-test it.