Student Research Projects

Spring 2016 Projects

Building a Negotiation Tool for Airbnb

Data from Airbnb was integrated with data from other sources, and by leveraging advanced machine learning models and data analytical tools, Airbnb customers gained insightful negotiation advice, helping them to get better deals at lower prices.

Restaurant Image Classification Using Deep Learning 

Images along with reviews are the most important sources of information for TripAdvisor's users.  In order to improve their website experience, TripAdvisor wanted to build a classifier for restaurant images.  To do so, we implemented convolutional neural networks, a machine learning algorithm inspired by biological neural networks, and reached an average accuracy of 87% on held-out test data.

Stochastic Query Optimization for Large Scale Text Search

Searching on social media about an entity or a concept can be challenging and time-consuming when such entities have ambiguous meanings.  This project, done in collaboration with Legendary Entertainment's Applied Analytics Division, provides an automatic searching query construction algorithm to help solve this problem.  The query-generating model was able to produce relevant tweets efficiently and ensure a reasonably low false positive rate.

Spring 2015 Projects


MBTA CAPSTONE - An attempt to improve the MBTA through Data.

The MBTA serves 4.8 million people throughout the Boston metro area and facilitates approximately 1.3 million trips each weekday. Aggregated entry and exit data is collected for each rail station at 15-minute intervals. Since commuting is one of the most habitual acts a metropolitan citizen performs, this data provides excellent means to predict ridership throughout the week.


Boston Globe Subscriber Conversion

The typical cyber-life of a BostonGlobe user starts with anonymous visits- from casually visiting the site, to ultimately becoming a subscriber. The Boston Globe would like to understand the idiosyncrasies and patterns of a subscriber and use that knowledge to increase subscriptions conversion rates.

Spring 2014 Projects


Driven Data: Data Science Competitions to Save the World

Isaac Slavit, Peter Bull | Advisor: Pavlos Protopapas

A website that runs kaggle-style competitions for non-profits. Non-profits have all sorts of technical issues and predictive questions that never get answered because they don't have the funds or the expertise to do the work.


Finch: A library for local search and stochastic optimization in Go.

Daniel Newman | Advisor: Pavlos Protopapas

Finch is a stochastic optimization library using Go 1.2, a language initially developed by Google. It includes methods such as Hill Climbing, Simulated Annealing etc. All of the functions were built to be as general as possible, so that they are useful for a wide variety of problems.


Structural Models With Optimization

Li Xiang | Advisor: Yaron Singer

A study on estimating structural models in general, optimization problems with ambiguous boundary conditions and random matrix theory. Many market equilibrium problems are modeled with these ambiguous boundary conditions in recent years and optimization techniques were developed to solve these problems.

Fall 2013 Projects


A Modifiable University Ranking System in D3

Connor Myhrvold | Advisor: Hanspeter Pfister

A university ranking program that re-weights and sorts universities based on user-input ranking component weights using QS / US News & World Report 2012 World Universities ranking list.


Parallel Cellular Dynamics Evaluation

Ryan King | Advisors: Lance Munn, Pavlos Protopapas

This project uses a CUDA to run real time simulation of tumor growth, allowing for accurate information about the size and extent of tumor growth that can help physicians save lives.

car dealer

Market Modeling and Computation Based on Random Utility Model

Muxi Li | Advisor: Professor David Parkes

Demand and supply analysis is a fundamental topic in Economics. By establishing proper mathematical models, one can have the flexibility to do market analysis including market share prediction, price elasticity estimation, new product introduction analysis, etc.