Hosted at: https://legoboy7.pythonanywhere.com/independent_project
I used the Julia programming language to ultimately create figures and animations for this project.
explain-maths.jlcreates two figures. One explaining linear functions, and the other, derivatives.hanoi.jlcreates the Tower of Hanoi problem and a search algorithm for solutions, both breadth-first search and A* with a few different heuristics. It then creates two animations, one showing the full solution and the other showing the first few steps of (breadth-first) search.hanoi-heuristic.jlimplements the "perfect heuristic" for Tower of Hanoi, which exactly determines the minimum number of moves (to the goal) from ANY position.icecream.jlfirst wrangles the data (found in "data" folder), then plots catplots of the features by month, then a scatterplot, then performs a linear regression which is added to the scatterplot. It also plots distributions.linear-regression.jlcreates some (unused) plots for understanding the analytical solution to linear regression. It also implements the gradient descent approach.nn-parameter-space.jlcreates and trains a basic neural network for demonstrating redundancy in extra layers with linear activation. It then implements my training solution for a "true" multilayer perceptron (i.e. using heaviside step function activation) that's just a logical extension of Rosenblatt's perceptron learning rule. It successfully solves the XOR problem.
The "figures" folder contains the figures generated from all these files, but also two figures I created myself using GIMP.
If you're looking to run these Julia files, then note they probably won't run as is— they were executed in a dynamic order using VSCode and the Julia extension.