/aho-ml-workshop-v21

Resources for AHO machine learning workshop spring 2021

Primary LanguageJavaScript

AHO ML / AI Workshop - Spring 2021

Code and links to relevant course materials for the AHO ML/AI workshop.

For the workshop we recommend using a modern browser, such as Chrome, Firefox or Safari.

Most of the exercises can be done in the p5.js editor, but it's also possible to download the code in this repository, edit and run the code examples locally on your computer.

For those of you who are interested in editing code locally on your computer, it's possible to use either Notepad (on Windows) or TextEdit (on Mac), but we recommend downloading a more advanced code editor such as VS Code or Sublime Text.

If you want to run the code examples in this repository, you'll need to set up a local webserver. You can do this by either using the Chrome extension Web Server for Chrome:

  • install the Chrome extension Web Server for Chrome
  • Start the extension
  • Click choose folder and point to the folder with these files.

or using Python (which should be installed by default on Mac) :

  • Open Terminal app
  • Browse to the folder with these files via cd command
  • run python -m SimpleHTTPServer 8887

After this, you should be able to go to http://localhost:8887 in your browser, and browse the files in this folder.

Day 1 : Intro to Machine Learning / classification

Intro to p5.js:

Basic p5.js examples:

ML5js examples in p5.js editor:

Teachable machine models in p5.js:

Examples of combining teachable machine models to do something in p5.js :

Bonus examples:

Resources:

Day 2 : Workshop, similarity and search, Generative ML

Search and Recommendations

Similarity maps

Sequential classifier models

Generative Machine Learning

Generative drawing:

Generative text:

Generative music/sound:

Generative Adversarial Networks

GAN examples:

Conditional GANs / Image translation:

Hands-on:

Collecting images

Potential sources of images :

Tools:

Day 3 : Training models, Multi-modal models

Controlling StyleGAN

CLIP:

CLIP-GLaSS (CLIP + StyleGAN) :

Machine Learning in the real world

Design resources: