This lab will give a quick example-based introduction to basic ideas in machine learning, using Python and scikit-learn.
We will also introduce elements of graph theory, network science, and the concept of patient similarity networks (PSN) using NetworkX - a Python package for the creation, manipulation, and study of the structure, dynamics, and functions of complex networks.
A. Lundervold & the Medical AI Assistant: Elements of graph theory and patient similarity networks (PSN) - A short introduction for ELMED219+BMED365 [PDF] [$\LaTeX$]
For medical and biomedical students new to graph theory, the following online tutorials and resources provide a gentle introduction:
-
Khan Academy Introduction to Graph Theory: A beginner-friendly series covering basic concepts and applications. Khan Academy Graph Theory
-
Graph Theory Tutorials by Sarada Herke: A YouTube playlist offering visual and intuitive explanations of graph theory concepts. Sarada Herke's Graph Theory Tutorials
-
Introduction to Graph Theory from University of Waterloo: Detailed course notes suitable for those seeking a more academic approach. University of Waterloo Graph Theory
❓ As Jupyter Notebook is quite new to many of you, you may want to skim through some tutorials. Here are two (also linked under "Getting Started" at MittUiB):
- https://jupyter-notebook-beginner-guide.readthedocs.io/en/latest/index.html
- https://www.datacamp.com/community/tutorials/tutorial-jupyter-notebook
| Notebook | 1-Click Notebook |
|---|---|
| Lab0-00-jupyter-notebook-markdown-basics.ipynb Jupyter notebooks and Markdown cells! In this notebook, we'll present non-coding cells in Jupyter notebooks to document and display information associated with coding cells. |
 |
| Lab0-01-simple-examples.ipynb Constructs predictive models based on some simple data sets. Provides a hands-on introduction to some basic ingredients and techniques in ML. |
|
| Lab0-02-networkx-tutorial.ipynb You can use NetworkX to construct and draw graphs that are undirected or directed, with weighted or unweighted edges. A large collection of functions to analyze graphs is available. This tutorial takes you through a few basic examples and exercises. |
|
| Lab0-03-patient-similarity-networks.ipynb Rather, the famous IRIS flower dataset with 4 different measurements from each of the 150 flowers - construction and exploring the "IRIS Flower Similarity Network". |
|
Spend some time playing around with the provided examples. You'll find some questions for you to investigate in the notebook. If you're already familiar with machine learning, you can try your hand at more advanced examples or, even better, help out other less experienced team members. Try out the things you learn in the DataCamp courses by modifying and extending the notebook used in this Lab.
If you have a subscription to ChatGPT Plus, you can also try out the the Medical AI Assistant (UiBmed - ELMED219 & BMED365) and see if you can get it to answer some of your questions.