- It’s an ongoing project of mine to develop a forecasting model for elections in Australia. I’ve made a number of posts on this website about the pitfalls and difficulties of coming up with such a model (starting here). I’m calling the model I’m building Emma Chisit; it wasn’t ready in time for the 2016 election but I’m hoping it will be for the next one.
- I wrote a primitive scraper for trading post website gumtree.com using the Python packages BeautifulSoup and requests, which persistently searches the site for a given query and sends an email using Mandrill when there are new results (for example when you’re after concert tickets).
- I wrote a trendy scraper for Google trends which gives daily data for queries across long timescales (Google trends only outputs weekly data for periods longer than a few months).
UPDATE (10/10/2016): Since I wrote this scraper, Google has changed its API and my code no longer works. I get a lot of requests about this code so hopefully one day soon I will find time to fix it.
UPDATE (18/02/2017): There’s a new version of the Trendy Scraper! See the post here, or see the updated code in the Github repo.
- I wrote a very basic poll aggregator (starting here) for the 2015 Israeli elections; like all the polls it aggregated it turned out to be very wrong in the end.
- There are some other posts on this website about data and data science:
- Here I quantify how much of a Pitchfork sychophant I am (not as much as I thought).
- Starting here I do some sentiment analysis of Twitter data for tweets posted during the broadcast of an Australian political panel show infamous for its lively Twitter feed.
- Here I wrote a tongue-in-cheek game to demonstrate the safety of air travel.
- In 2016 I started looking into simulating cricket matches with the goal of performing some data mining on the simulations; the series of posts starts here.
- My PhD thesis, submitted in 2014, was titled Alternating quiver Hecke algebras. It consisted of a careful and complete study of a new family of algebras, which generalise and build on important recent work. The main theorem of my thesis gives an explicit isomorphism between my new family of algebras and a family of classical objects which have been studied for over a century, providing new insight and structure for these objects. I wrote a FAQ about my PhD here.
- You can find a list of my published mathematics papers here.
- I was involved in creating a series of audio-visual learning modules for the University of Sydney’s Mathematics Learning Centre to help HSC students quickly develop mathematics skills which are prerequisites for university courses.
- I ran the first-year linear algebra course MATH1002 for the University of Sydney Summer School in 2015. I took a much different approach to the course than in previous years, running the whole thing through a Wordpress blog. I received a lot of very positive feedback from students on this experimental and forward-thinking course design.
- I’ve been working on the John Hopkins University Coursera Data Science specialization, which consists of nine courses in various aspects of data science (although I’m not working to their timetable or paying for the courses).
- I completed the Stanford University Algorithms Part 1 Coursera course and am currently working on Algorithms Part 2.
- I completed the Coursera specialisation on Functional Programming in Scala.
- I worked on a Kaggle competition for learning social circles from Facebook data and have written some basic models using the Python package networkx for thinking about the data and the problem.
- I’m a keen amateur (very) photographer: all over this website you can find examples of my photographs taken all over the world.
- Here I made a simple Python application which reminds busy expats to call their mums.
- I very much enjoy jogging, particularly in new and unexplored places.
- When I was twelve I wrote a novel about a motley crew of magical futuristic cat detectives. It was well-received by my mother: I’m still searching for a publisher.