I love markets, numbers, and the natural world

and this is a small collection of things I like.

I'm a trader, a manager, a builder and a dad. I've been a HFT guy for over a decade, trading all sorts of derivatives in Korea, Japan, Hong Kong, Singapore and Australia. When I'm not trading, you'll probably find me crunching some numbers, or building and tuning some Machine Learning to do the crunching for me.

If you're like me and get excited by Machine Learning and Data Visualisation, there are a selection of fun projects listed below.

The island of Tasmania is ~42% protected land, so it was easy to develop a passion for nature growing up there. Many weekends during my childhood and teenage years were spent camping in the parks and reserves of Tasmania, and while I know I complained about the cold and snow, it was an upbringing I was fortunate to have.

I got my first camera at the age of 10, a hand-me-down, back in the age of real film, and I've been documenting the landscapes, wildlife and people I encounter ever since.

All images are my own work - No shutterstock here. Mouse over a photo for a description.

Machine Learning

I love machine learning and neural nets, and enjoy tossing them at varied problems to generate interesting solutions.
Sometimes I build them just for fun, like my Convolutional Neural Net Baby Monitor which watches the babycam to ensure the little guy doesn't kick his blankets off.

Here's a fully interactive demo of some data clustering using a Support Vector Machine.
I know SVM's are a little old hat, but they have a place, and more importantly are fast enough to train that you can run them interactively in an online demo like this.

Fake News

Fake news is everywhere, which motivated to me to build a machine learning app to decide if a news article is real or fake.
I also wanted to explore the use of efficient online algorithms and Natural Language Processing, so this app combines both those elements, using a Passive Aggressive Classifier as its learner, and TFIDF for word vectorisation.

Liquidity Measurement

This is the live graph of a liquidity tracking algo I designed for optimising trade sizes based on market conditions
It's using some historical data for this demo, rather than a live market feed, but it shows the relevant bid/ask liquidity dropping off before large moves, and a clear negative correlation between bid and ask liquidity.

I'm progressively adding further visualisations and features to the site, so check back sometime to see what's new.

The Natural World

A few of my favourite shots from a recent journey to South America.