diana lam

MTALead - Modeling Pedestrian Patterns During the L Train Shutdown

The April 2019 L train shutdown in New York City will impact 225,000 people per day. How can we collect data on pedestrian movement patterns and estimate how they will change with the shutdown? In this post I describe MTALead, a proposal to collect data on and model real-time pedestrian flows to mitigate the impacts of the shutdown on riders and neighborhood business owners.

Perceptions of Urban Perception - Perceived vs. Actual Safety in NYC Neighborhoods

How accurate is our perception of streetscapes in determining the actual safety of urban neighborhoods? In this project I identify neighborhoods in New York City that are perceived as more or less dangerous than their crime rates would indicate, and analyze the socioeconomic, demographic, and physical features that characterize these neighborhoods.

Investigating the Twitter Strategies of the 2016 Presidential Candidates

It's election year and social media is playing a larger role than ever before in facilitating interaction between the 2016 presidential candidates and the general public. How are the candidates leveraging their Twitter presences at this stage in the race and how does this compare across party lines and amongst candidates? In this post I look at topics, sentiment, and various "call to action" strategies of Clinton, Cruz, Rubio, Sanders, and Trump.

Predicting Yelp's Elite

How a Yelp user's activity levels, popularity, review sentiment/content, and review structure impacts his or her chances of being a Yelp "elite." This is the second and concluding part to the exploratory insights discussed earlier.

Big Data and Smart Cities

By 2050, 75 percent of the world's population will live in cities. How can data and technology optimize city services and improve quality of life?

Exploring Yelp's Elite

What differentiates Yelp's elite users from the non-elite? In this post I do some initial exploratory data analysis to isolate the key features that begin to answer this question.

Predicting the Success of Lower Budget Movies - An Analysis of Revenues vs. Recognition

Using linear regression, can we isolate the features that drive the success of movies with budgets under $30 mllion? How are these features different for success as defined by revenues versus recognition (in terms of Oscar nominations)?

Hands-off Scraping with the Twilio REST API

Recently, I've been working on a project that involves scraping data for over 16,100 movies. Rather than sit and watch the terminal output of the script, I set up a text alert to let me know when it's done.

Munging the MTA - Optimizing a Non-Profit's Street Team Placements

Outside what subway stations should a women's tech non-profit deploy their street team members in order to drive attendance to their annual gala and spread awareness for their cause?