Jonathan Mackenzie

Phd Student, Flinders University
Android. Python. Fencing.

About Me

I am a PhD student at Flinders University. My research focuses on predicting and detecting incidents in short term traffic flow.
Read my thesis here
My other interests include self-teaching and learning about:

  • Algorithmic art
  • Machine learning
  • Parallel programming with OpenCL
  • Web development with a Python/Linux/nginx/bootstrap workflow
  • Veganism
  • Coaching fencing


Most of my projects, including my PhD work and publications, are available open source on github profile

  • Rpi-can-logger project: software package for raspberry pi to log, interpret and upload CAN data from a variety of vehicle types
  • Webcan: a web based viewer and analysis suite for use with rpi-can-logger
  • nn-cl: an in-development neural network library written in opencl/pyopencl/python
  • HTM-Models-Adelaide: project for predicting traffic flow and visualising results
  • img_term: a fun library to render images/video/usb camera to your terminal
  • pyramid_mongodb2: plugin for pyramid to better integrate mongodb into your pyramid project, including a comprehensive debugtoolbar addon
  • traffic-prediction: using LSTM with python/keras to predict traffic flow
  • TonsleyLEDManager: pyramid/mysql/pygame project to display animations on the LED display at Flinders Tonsley
  • Bibtex-Label-Creator: generate Bibtex labels for Endnote
  • Markov-img-gen: Experiment in learning/generating images as a Markov process using maximum likelihood
  • Album2mp4: script using python/ffmpeg to generate videos from albums
  • Twatbot2: IRC bot with a lot of features
  • Android Oscilloscope view: a custom view for Android that renders a view as an oscilloscope would with an erasor
  • htm-over-http: a small library to run and interact with HTM models via REST interface

Open Source Contributions

I've contributed to a number of open source projects, including:

  • Nupic: a biologically inspired framework for prediction and anomaly detection
  • Hyperas: a template based library for hyperparameter optimisation

Android Apps

Released a number of apps on the google play store for Android:


Traffic Prediction using HTM and LSTM

An Evaluation of HTM and LSTM for Short-Term Arterial Traffic Flow Prediction: Published in IEEE Transactions on Intelligent Transportation Systems, investigating and comparing HTM and LSTM for short term arterial traffic prediction


Feel free to contact me through the following means: