Recommended learning sequence (all free):




Basic skills:


1. Python (5 hours) – Basics of Programming


2. Pandas (4 hours) – Data processing


3. Data Visualization (4 hours) – Data Visualization




Machine learning:


4. Intro to Machine Learning (3 hours) - Introduction to ML


5. Intermediate Machine Learning (4 hours) - Advanced ML


6. Feature Engineering (5 hours) - Feature Engineering




Deep learning:


7. Intro to Deep Learning (6 hours) - Introduction to deep learning


8. Computer Vision (4 hours) – Computer Vision


9. NLP (3 hours) - Natural Language Processing

Steps to get started with Kaggle competitions:

1. Start with the "Getting Started" contest
   - Titanic (classic classification problem)
   - House Prices (regression problem)
   - Digit Recognizer (image classification)

2. View Notebooks to learn other people’s solutions
   - Click on the "Code" tab
   - Sort by number of votes
   - Learn the ideas of high score scheme

3. Submit your own predictions
   - Fork a basic Notebook
   - Modify and submit
   - View ranking

Steps:


1. Log in to Kaggle and create a new Notebook


2. Click "Settings" on the right


3. Select "GPU T4 x2" in "Accelerator"


4. Start running the deep learning code




Free quota:


- 30 hours of GPU time per week


- Supports T4 GPU (16GB video memory)


-Support TPU (in some cases)




Suitable for running:


- Image classification model training


- Small LLM fine-tuning


- Kaggle competition code