Beniamin BOGOSEL

Introduction to Machine Learning using PyTorch

Statistical learning

Course 1

• Notes Course 1: pdf
• Notebook Regression 1D Google Colab: link
• Linear regression: Notebook
• Lab pdf
• Notebook Lab notebook
• Solution Lab notebook

Course 2

• Linear regression: pdf
• Notebook A: notebook
• Notebook B: notebook
• Notebook C: notebook
• Notebook Additional Work: notebook

Neural Networks

In this course we will explore facts about neural networks: definitions, characteristics, training, etc. The practical aspects will be done in Python using the PyTorch library.

Course materials are taken from https://github.com/mrdbourke/pytorch-deep-learning. We will work directly using Jupyter Notebooks exploring practical aspects regarding the workflow in PyTorch.

Course Materials

1. Pytorch basics: Jupyter Notebook: file course 1. Notes for first course: pdf file.
2. Pytorch workflow, Linear Regression: Jupyter Notebook: file course 2. Notes course 2: pdf file.
3. Classification, Non-linear activation: Jupyter Notebook: file course 3. Notes course 3: pdf file.

   Exercises: file exercises

   Example 1D: file exercises

4. Computer Vision: convolutional networks Jupyter Notebook: file courses 4-5. Notes Course 4: pdf file.
   • Exercise file: Notebook
   • Solution file: Notebook
5. Custom Datasets: Notebook
6. Using existing models/arbitrary size images Notebook
   • Model weights: resnet18.pth (Google Drive). Download the file and save it beside the notebook.
   • Link to zip file: zip file
7. Transfer learning: Notebook

   Transfer learning exercise: Notebook

8. Experiment tracking: Notebook

   Experiment tracking exercise: Notebook (Google Drive)

Project Subjects: pdf file

Notebook for generating 2D data sets: Notebook.