Last updated: 2024-01-17 (with bmed365 kernel), A. Lundervold
This lab is part of our journey through the concepts and applications of deep learning in medicine and biomedical research.
Some of the material has inspiration from and links to the work of great educators and researchers in the field, e.g. Grant Sanderson, Daniel Bourke, and Andrej Karpathy
If you have a subscription to ChatGPT Plus, you can also try out the the Medical AI Assistant (UiBmed - ELMED219 & BMED365)
GPT and see if you can get it to answer some of your questions. See also Q&A-in-the-wild
<in progress ...>
(in the order of duration ...)
-
What is Deep Learning? (Deep Learning Tutorial For Beginners, 2023) by Simplilearn - [link] (5:51 min)
-
But what is a neural network? by Grant Sanderson - [link] (18:39 min)
- See also his 3Blue1Brown: The basics of neural networks, and the math behind how they learn - https://www.3blue1brown.com/lessons/neural-networks - code
- and: Gradient descent, how neural networks learn - [link] (20:33 min)
- and: Analyzing our neural network - [link]
- and: But what is a convolution? [link] (23:00 min)
- and: What is backpropagation really doing? [link] (12:46 min)
- and: Backpropagation calculus [link] (10:17 min)
- all made with his Manim engine - see his TEDxBerkeley talk ... (and also the ManimCE Example Gallery)
-
Building a neural network FROM SCRATCH (no Tensorflow/Pytorch, just numpy & math) by Samson Zhang - [link] (31:27 min)
-
Deep Learning State of the Art (anno 2019) - MIT by Lex Fridman - [link] (46:24 min)
-
MIT Introduction to Deep Learning 6.S191: Lecture 1 in Foundations of Deep Learning by Alexander Amini - [link] (58:11 min)
- For all lectures, slides, and lab materials in MIT 6.S191 see http://introtodeeplearning.com and https://github.com/aamini/introtodeeplearning
-
The spelled-out intro to neural networks and backpropagation: building micrograd by Andrej Karpathy - [link] (2:21:51 hr)
- "It only assumes basic knowledge of Python and a vague recollection of calculus from high school ..."
- micrograd on github: https://github.com/karpathy/micrograd
- Jupyter notebooks built in the video: https://github.com/karpathy/nn-zero-to-hero/tree/master/lectures/micrograd
-
Deep Learning Crash Course 2023 | Learn Deep Learning Fundamentals In 5 Hours | Simplilearn - [link] (5:23:36 hr)
-
Learn PyTorch for deep learning in a day. Literally. by Daniel Bourke - [link] 5 Chapters (25 hrs, 36 min and 57 sec long)
- For data, code, exercises, and slides - https://github.com/mrdbourke/pytorch-deep-learning
- Read the course materials online - https://learnpytorch.io
-
Convolutional Neural Networks for Visual Recognition the CS231n course @ Stanford University by Fei-Fei Li et al. [link] (10-week course)
- GitHub repo: https://github.com/cs231n/cs231n.github.io
- Python Numpy Tutorial (with Jupyter and Colab): https://cs231n.github.io/python-numpy-tutorial
You can read the Lab2-DL materials on any device but this course is best viewed and coded along within a desktop browser.
For the hands-on Lab, Google Colab can be used. If you have no experience with it, go through the free Introduction to Google Colab tutorial.
To dig into Pytorch and tensors:
- Click on link "00. PyTorch Fundamentals"
- Click the "Open in Colab" button up the top
- Press SHIFT+Enter a few times and see what happens