Tutorials and followtorials for the course "Machine Learning: A Hands-On Approach" at University of Bayreuth by Christopher Kuenneth

Run the notebooks on Google Colab using the buttons in the notebooks or in VScode or Jupyter.

I am using PDM for managing the Python dependencies.

Lecture 1

1. Find out how good you are at Python Click
2. Deploy the 13-billion CodeLlama Instruct model on an Colab instance. Click

Lecture 2

1. Python ABCs Click
2. Python XYZs Click
3. A first AutoML ML pipeline Click

Lecture 3

1. Finish: A first AutoML ML pipeline Click
2. Intro to data processing Click

Lecture 4

1. AT HOME: Work with Pandas: balance data, plot, store Click
2. Outlier detection with sklearn Click
3. Compute fingerprints for text and materials Click

Lecture 5

1. AT HOME: Operations 🤕 on multi-dimensional arrays (tensors) Click
2. Step-by-step linear regression Click (moved to lecture 6)

Lecture 6

1. Template for your own project Click
2. Step-by-step linear regression Click
3. HOMEWORK Click

Lecture 7

1. Step-by-step linear regression Click
2. My first NN Click
3. HOMEWORK (play with and modify some parts of step-by-step linear regression) Click

Lecture 8

1. AutoML with AutoGluon to predict the tendency to crystalize for polymer Click
2. HOMEWORK: prepare for your project 🚀 Click
3. Visualize neural networks Click

Hackathon 1

1. Regression: template for training a property ML predictor from chemical structures Click
2. Computer vision: template for training a property predictor from figures Click

Lecture 9

1. Train your pytorch NN Click
2. HOMEWORK: Read https://jalammar.github.io/illustrated-transformer !

Lecture 10

1. Please read this great article on "A Visual Exploration of Gaussian Processes"
2. Folloturial: BO for design of experiments Click

Lecture 11

1. RAG with Wiki data for polymers Click
2. Deploy a ML model with streamlit Click

Hackathon II

See README in folder.