Data Science and Artificial Intelligence I

Johann Mitloehner, 2024

• Symbolic AI
  • Answer Set Programming with Clingo
  • More Examples
• Neural Networks
  • pre-reading Tariq Rashid, Make Your Own Neural Network, CreateSpace, 2016
  • Neural Net Introduction: One-layer Feed-Forward ||| nb
  • Examples: Single-Layer
• Extending the basic model
  • Two-layer Net and Backpropagation ||| nb
  • More Examples: NN with mushrooms and sentiment sentences
• Connectionist Machine Learning Framework: PyTorch
  • Adaptive models with PyTorch ||| nb
• Deep learning, word embeddings, and business applications
  • Embeddings, LSTM ||| nb
  • Comparison: Decision Tree
• Transformers and Large Language Models
  • Transformers, GPT, BERT ||| nb
  • Build your own Transformer ||| nb
  • Post-module assignment:
    • business application problem solved in connectionist machine learning
    • dataset e.g. from UCI repository
    • submit via Canvas

We only use free Open Source software, so you can install everything on your own computer.

• If needed, get Python from python.org and click 'Add to Path' when you run the installer.
• Use pip on the command line to install additional packages as needed, e.g.
  pip install jupyter pandas matplotlib scikit-learn

However, we also have Jupyter notebook servers ready for you at the Canvas web page for this lecture. You can use the server via your web browser.

More Examples:

• Dec 1, 2023: UCI Wine Dataset
• Dec 22, 2023: UCI Student Performance

Older videos:

• Neural Net Intro
• Jupyter Notebook
• Multi-layer and Backprop
• Simple RNN, LSTM, Decision Tree
• DT v2