Machine Learning¶
Some ML content I’ve made, some of which I use for self reference.
Builds
Semi-Supervised Learning Using Virtual Adversarial Training, this was an exploration in what is the furthest you can go on few labels.
Modeling ML methods using regression for sin fitting and digit classification, RL for simulated pole balancing, and recurrent NN’s for language identification.
Autocomplete from scratch.
Feature detection using Harris key points, project summary here.
Bonus: first ever CS project, stock prediction by parsing 10K’s using BeautifulSoup.
Notebooks
These are extensive Jupyter notebooks I’ve written for ML textbooks.
Interactive Animations
Building these helps me solidify concepts, this selection is ML related.
Most are interactive in some way and likely don’t display well on mobile.
Collapsing cells as you go eases rendering.
Math
Derivative
Limit
Min, Max
Integral
Bayes
Button press simulates finger position on catching a globe.
Taylor Series (interactive sliders)
Machine Learning Concepts
Linear Regression
Click.
Least Squares
Click and drag.
Perceptron
Click, drag and spacebar.
K-Means