Programming in Python for Machine Learning

Hi there! This is an interactive tutorial on how to program in Python with an additional focus on Machine Learning (ML). I use it to teach Python for ML, i.a. at the Artificial Intelligence study at the Kepler University Linz.

What awaits you here?

• A full tutorial on how to program in Python from the beginning (bits/bytes) to programming neural networks in PyTorch.
• Interactive materials for self-study.
• Small tasks for you to check your knowledge.

Which areas are covered?

• Basics about computers and programming (CPU, GPU, RAM, harddrive, bits, bytes, datatypes)
• Set-up of Python and PyCharm.
• Basics in Python (variables, datatypes, lists, tuples, dictionaries, functions, modules, exceptions, files)
• More advanced Python (subprocesses and multiprocessing, regex, numpy, matplotlib, classes, code optimization, numba, PyTorch)
• Python for Machine Learning (PyTorch)

Requirements

• No prior knowledge in programming is required. If you're already familiar with similar languages or Python, just skim over the first units.
• Laptop, PC, or access to a server is required. No high-end machines are required. GPU is optional (but faster and more fun).
• 64bit Python 3.6 or higher. (For installation instructions see slides.)
• Recommended operation system: Ubuntu 18.04 or higher.

Structure

• Materials are structured in Units. Units typically consist of (interactive) Python code demonstrations/explanations, followed by small tasks to test your knowledge and gain experience.
• Two parts:
  1. General Python programming: basics up to more advanced things, including PyTorch preview
  2. Python and Machine Learning: Building ML projects in Python using PyTorch, following ML standards and good-practice, some hints for neural network training (will be uploaded later)

How to start

• Start with part I in folder "Programming-in-Python-I" and follow the README.md file there.

Best wishes and have fun!

-- Michael Widrich (widi)