I’m Matt. Above is a picture of Mount Fuji that I
captured during my travels through Japan years ago.
Mount Fuji has been the subject of many great works of art
(Hokusai’s Thirty-six Views of Mount Fuji,
for example).
In many ways it is the ideal subject, infinitely patient, beautiful,
static against a constantly fluctuating backdrop of weather and seasons.
When I’m not admiring nature and art, I am working with data.
In some ways analyzing data is like drawing Mount Fuji. A data analyst
recognizes a subject (structured collection of data points)
behind clouds of interference (noise)
and approaches that subject with a new way of seeing (statistical tools)
to extract meaning (insights).
An insight is a nugget of a bigger idea, from which the artist
delivers value in the form of an evocative picture,
and the analyst delivers value through
visualizations, reports, recommendations, and statistical models.
If you can’t tell, I have a deep fondness for data,
and in particular for statistics and machine learning, which are the tools that
guide me through the mist toward structure and meaning.
I find it especially thrilling to implement statistical ideas through
code and neat data analyses.
I like to write about my favourite projects and discoveries.
I’m sharing some of that here and here.
I also make music on the side – you can find that here.
This site is under renovation.
Many more articles and reports coming soon.
In the meantime, view a site overview below.
▼ Software
► datopy [docs]
- A Python package for working with unstructured data, providing a simple interface
for web scraping, data modeling, and building ETL pipelines.
Tools used: Pandas, Pydantic, PyTest, MyPy, Sphinx, GitHub Actions, Seaborn, Beautiful Soup
► autocv [docs]
- An R package automating repetitive elements of the job application process,
including managing work history data, building and rendering documents, and
performing text analysis to extract and evaluate keywords.
Tools used: dplyr, stringr, purrr, pkgdown, Excel, LaTeX, CSS
▼ Resources
► Machine learning tour
- A high-level tour of the theory and implementation of machine learning
algorithms in Python.
Featured: Supervised learning in Python (draft)
► References & docs
- An assortment of tools for working with data, including syntax
reference sheets, high-level outlines, and checklists.
Featured: Technical reports with R Markdown & Bookdown (draft)
► Demos
- A short series of demos using a single dataset to illustrate the
application, strengths, and weaknesses of a variety of machine learning
algorithms.
Featured: (In development)
▼ Analysis
- A place to share my findings and process for analyses and models
that have yielded interesting insights.
Featured: (Do changing attitudes toward the environment reflect accelerating extinction rates?) (In development)
▼ Blog
- A home for my musings on math
(proofs and derivations of my favourite results and algorithms),
machine learning (implementations and simulations), and
everything else (reflections on music, film, society, etc.).
Featured: Bias-variance decomposition (In development)
▼ Music
- An ongoing catalogue of some music I’ve made.