Notes on Davis Rippon

Ant Colony Optimisation

Wed, 15 Jul 2026 08:24:08 +0000

The following is a article adaptation of a research project I did on Ant Colony Optimisation. See the code that I wrote to do visualisations like this:

Your browser does not support the video tag.

1. Introduction

The Travelling Salesman Problem (TSP) is an NP-hard combinatorial optimisation problem aimed at finding the minimum-cost Hamiltonian circuit¹ for some weighted graph.

TSP is ubiquitous across various domains and fields, such as network performance, package delivery routing, water resource management, and beyond [1].

Motion Compensation in Video Codecs

Fri, 10 Jul 2026 13:57:13 +0000

Why is it so “cheap” to stream video, bandwidth-wise? If we look at it naïvely, we may think a 1080p video @ 30fps with three colour channels should stream at a rate of vaguely something like:

In 2026, 186.6 Megabytes per second is still greater than most people’s internet speed altogether. Considering that we’ve been able to stream video for decades, well before internet speeds were in the order of Megabytes, this indicates some other tricks are taking place to allow video to happen.

Typst Showcase

Thu, 25 Jun 2026 12:29:41 +0000

I can now write articles in Typst thanks to George Honeywood’s Blog. Go read his article and clone his Hugo fork if you would like to do the same!

Math

Tables

	Exam 1	Exam 2	Exam 3
John		A
Mary		A	A
Robert	B	A	B

Diagrams

Figure 1: ant at

𝑘_{𝑖}

chooses next node, with

𝜏_{𝑖 𝑗}

in purple and

𝑑_{𝑖 𝑗}

in green

Making a library-less MLP

Tue, 14 Jan 2025 14:24:39 +1100

I’ve made a Multi-Layer Perceptron (MLP) in C++ as a learning project. The goal, initially, was to make a replica of the MLP described in 3Blue1Brown’s lectures, with no reliance on ML libraries.

What I’ve made achieves that. It trains from the MNIST database for 25 epochs/until a loss threshold is exceeded and then predicts some numbers from the test set correctly.

It’s Not that Good

As of right now, it is quite bad. Size is kept small because I’m running it on a laptop, all of the matrix multiplications are implemented with for loops and nothing is parallelised.