Reading List
Data Science and Cultural Objects
Recent TDS articles that go above and beyond
As mentioned in my introduction piece, I am always on the lookout for TDS articles that connect the world of data science to some cultural artifact or trend, from consistencies in 19th century novels to a city’s public transportation network. While I find straightforward tutorials or speculations on the future of AI quite useful, I am always immediately drawn to pieces that make a connection to the “real world.” Below I have listed TDS pieces from the past two weeks that stood out to me especially.
The Perils of Palette Transfer by Syafiq Kamarul Azman
This long-form piece offers up an understanding of and tutorial for the digital replication of colours, demonstrating how to do so using a photo of a peony. The author elaborates a bit on this, but what I found intriguing was how a machine can so readily and easily identify and duplicate colour spaces, which are so vast and still very much unknown.
Create a Spotify Playlist for yourself and your partner or friend by Merlin Schäfer
Here, the author reflects on Spotify’s Duo Mix subscription service aimed at combining two users’ music tastes. Shäfer and his partner did not find the feature to work ~just right,~ so he made one himself! This piece serves as an easy-to-follow tutorial on how to recreate Duo Mix, but just how he and his partner wanted.
Can foods help us fight COVID-19? by Guadalupe Gonzalez
We know that getting enough vitamin C in our bodies on a regular basis can boost our immune systems, but what should we be eating during a pandemic? This piece, based off of a Human Genomics paper, “applies graph machine learning to predict molecules within foods with anti-COVID-19 properties based on their capacity to target the SARS-CoV-2-host gene-gene (protein-protein) interactome.” The results? Antiviral hyperfoods include different berries (blackcurrant, cranberry and blueberry), cruciferous vegetables (cabbage, broccoli), apples, citrus fruits (sweet orange and lemon), onions, garlic and beans. So stock up!
Reverse Geocoding with NYC Bike Share Data by Clif Kranish
Biking around New York City isn’t the pleasant experience that most people imagine, but this author seeks to improve the logistical experience, at least, by adding the borough, neighbourhood, and zip code where each station is located. The tutorial is followed by a short analysis of where bicycle trips tend to start and end, which is always good to know for traffic reasons, if anything.
A Mixed-Integer Optimization Approach to Rebalancing a Bike-Sharing System by Duncan Wang
More bike content! BIXI Montreal, which is the original urban bike-sharing program, exists as a form of public transportation on the island during the spring, summer, and fall. Wang seeks to solve the bike imbalance problem (wherein many bikes end up in popular locations in the city and some stations are left empty) for BIXI by using mixed-integer optimization. The tutorial was thorough and in the end, provided a theoretical solution to the problem at hand. As a “Bicycle-Oriented Teen” and Montrealer, this was a very entertaining and enlightening read.
Heated Discussions: Predicting Conflict Intensity Using Climate Data by Richard Pelgrim
Pelgrim discusses the correlation between climate change and increased global armed violence in this his piece. He uses a plethora of graphs and maps to document where conflict occurs, how often, and change over time. Seeing as climate change is and has been a social problem, this article brought a particular issue to light regarding results: More research efforts should exist in the parts of the world that see the most conflict.
Tweet Analysis of #gamestop During the Stock Saga by Skanda Vivek
In 2018, it was my job the effects that current events, filtered through Twitter, have on the stock market, but I never fully quantified the importance of this medium. Here, this author examines the direct impact that #gamestop tweets had on the stock market from January 25–29 and provides interesting conclusions and unanswered questions from his data.
Democratizing Historical Weather Data Analysis with R by Sabi Horvat
As a plant hoarder, this piece instantly caught my eye. This article provides a very in-depth tutorial on how to determine the best plants and crops for a given area based on five different factors and their corresponding maps. These stats are calculated by postal code, so there is something for everyone to learn.
Until next time, happy reading!
