Striding through the Boston Marathon Winners dataset from start to finish line — As I stood amidst the electrifying atmosphere of this year’s Boston Marathon, cheering on the runners with exhilaration, I began to understand the almost inarticulable magic of Boston. Marathons demonstrate that with mental fortitude, discipline, and determination, ordinary individuals can accomplish extraordinary feats. Even larger than that, marathons are a…

Evaluating different heuristics to determine the most effective solving strategy and building an AI-powered assistant tool to help you win — Wordle has reignited group chats, enlivened everyone’s competitive spirits, and inundated Twitter with cryptic emoji grids since it achieved viral status in January of 2022. …

Digesting the Digest: Reverse-engineering my Medium Interests with Topic Modeling (Part 3) — This marks the last article of a three-part investigation to better understand my Medium suggested reads by way of an NLP technique called topic modeling. In this article, I will introduce what I believe to be the most powerful topic modeling algorithm in the field today: BERTopic. I will also…

Digesting the Digest: Reverse-engineering my Medium Interests with Topic Modeling (Part 2) — Motivating Topic Modeling In my last article, I walked through how to extract metadata from email newsletters. Specifically, I demonstrated how to use the Gmail API to collect Title, Subtitle, Author, Publication, and Minutes (an estimate of article length in reading time) for each suggested article in my Medium Daily Digest newsletter. With…

Digesting the Digest: Reverse-engineering my Medium Interests with Topic Modeling (Part 1) — I’ll be honest. I am obsessed with the way in which Medium’s recommender system oftentimes knows me better than I know myself. Their suggested reads are so enticing that I wish I could wake up each morning and devour 10–11 articles that are perfectly in line with my interests. That…

Creating relational plots, categorical plots, and GitHub-style heatmaps with my Peloton workout data — Motivation Many of us will remember the day the COVID-19 pandemic was declared. It is safe to say that this monumental event represented a major upheaval in our lifestyles and routines. Many took to sourdough baking immediately, some grew addicted to watching Tik Tok videos, and others — more power to…

Querying a PostgreSQL Database with Python, displaying results with Pandas, and visualizing with Matplotlib — Now more than ever, I am wishing I was a professional athlete in the best shape of my life, gearing up for the Tokyo 2021 Olympics. But rather than feeling sorry for myself as I set up my spot on the couch with pillows, blankets, and a non-vegetable snack, I…

Using the TextBlob API and word cloud visualizations to synthesize the discourse surrounding global vaccination efforts and progress — It’s been a long year of sickness, devastation, grief, and hopelessness, but the global rollout of COVID-19 vaccines has sparked feelings of relief and newfound optimism for so many. The discussion of vaccination progress, accessibility, efficacy, and side effects is ongoing, and it is permeating through news stories and Twitter…

How I repurposed my K-Means Clustering and PCA code into a sleek web app for analyzing Spotify playlists — In the age of information, beautiful interfaces are not only admired, but they are expected. Spotify Wrapped is an annual campaign which serves the purpose of bringing all Spotify users personalized insights for their soundtrack of life. It is fully powered by big data and analytics, yet what promotes it…