Grant Holtes

Exploring the world through the lens of statistics, digital technology and economics. Currently working in Analytics and AI consulting at Deloitte in Melbourne, Australia and studying Computer Science at the University of Melbourne.

Some things I've done: LinkedIn Resume

Technical Projects

Projects mainly in data science, econometrics and software engineering.

[2022] Flow programming for computer vision applications

[2022] In browser TensorFlow object detection demo

[2021] Playlist optimisation for Spotify

[2021] Mapping Melbourne's cycling infrastructure with OpenLayers

[2021] Extracting Colour Palettes with Unsupervised Learning

[2021] Agent based modelling exploration

[2020] Statistical analysis of service deserts in Victoria: report, code

[2020] SortStream ML powered document sorting

[2020] DocDump text and document metadata extraction

[2020] Automated data exploration in R: code: Titanic example Economic example

[2019] Unsupervised ML image embedding to measure art style and visual similarity

[2019] Dimensionality reduction for art style embeddings

[2019] Navigation App for Urban Cyclists: Write up Code "The Submarine" feature

[2019] Exploration of ML Document Classification Techniques - Code

[2019] Interacting agent simulation and statistical physics to model fashion epidemics

[2019] Deep Dream implementation in Keras

[2018] A replication and extension of the paper "The Impact of Economic Conditions on Participation in Disability Programs: Evidence from the Coal Boom and Bust"

[2018] Economic and financial market evaluation with similarity methods

[2018] How Animal Investors Beat the Market

[2018] URL Sketchifier: Give your urls a look that Gmail will flag as suspicious

[2018] Logistic Regression — Understanding Explainable AI: write up code

[2018] Decision Trees — Understanding Explainable AI: write up code

[2018] Building Google's art and culture portrait matcher

[2018] Analysing bird flight by merging videos in MATLAB

[2017] Developing a compass web app with mobile sensor integration

[2017] Developing autoencoders for image correction

[2017] Mini-max algorithm for Connect 4

Creative Projects

Projects in generative art, image manipulation.

[2022] Weighted attractor rendering

[2022] Rendering attractors at 60 FPS

[2022] Organic growth emulation in TouchDesigner

[2022] Esher's Cube Space Division in TouchDesigner

[2022] Creating sunset gradients in P5.js

[2022] Creating abstract images from lines

[2021] Abstract visualisation of what a state of the art object detection model is seeing