Yelp Review Generater

Created an application that generated fake yelp reviews from a string. Written using a recurrent neural network (RNN) trained on the Yelp public data set. This was created as part of a Data mining course at WPI. 

Note: there was VERY little time to train this RNN. As a result this fake reviews are more entertaining than realistic. 

Generating with seed: " once with an associate, or who I though"

once with an associate, or who I though snain.
I have that when aurnt for our ple only ove Anly greever's 
proaptan't chownit ware ses with we't greetery is finausys 
wi were gire dy haved their andtears. Deintesbesse efled. 
nously wish and the price...the appievery ok and free Gordy 
and with Cheep for you was share so right some place but in wark as,
 reggesting other every pubar and it at werom, an everyther of luderar-owcage the pooly,    

Project Pather: A Faulkner Hospital Mapping System

Served as the lead software engineer for a team of 10 students creating a mapping system for Faulkner Hospital. This application would allow patients to search for doctors and see how to get from a kiosk to that location (think google maps for inside a hospital). 

This project was created entirely in java. The front end was created in JavaFx. the back end was created in Java and interfaced with a Derby SQL database. The path finding was done with the A* algorithm. 

As lead software engineer I was responsible for creating the design for the system in UML and helping the less experienced developers on the project. 

The code can be found on github


A simulation of a virus breaking out in a college campus. This is done by creating an artificial schedule for each virtual student and having them go to classes. Each sick student has a chance to pass on their sickness to their class and spread the virus. This part was done using the MASON library. The path of the transmission of the virus was then stored as an adjacency matrix and visualized using Gephi's visualization software. 


Riot Games API Analysis

Data analysis of RIOT's online game league of legends using their API and written in python. Data on each online game is stored by the RIOT servers and is publicly accessible through their API.  Using this API, I attempted to correlate the relative performance of players based on how much of an advantage they had in the game. This involved parsing each JSON file containing the match data and finding the data I wanted. Then the Python library matplotlib was used to create linear regressions to see how strongly performance correlated with an in-game advantage. 


A visual representation of Quick Sort, Merge Sort, and Heap sort. 
Numbers correspond to a height and a color in a pink gradient. 
Code written in Java. Graphics Drawn using Java Swing.


A C++ implementation of the classic snake game. Uses the graphics library SFML for timers and graphics. 
            Git Repository: 


The arcade style breakout game written in Java. Uses the swing library for graphics and timers 
    Git Repository:

(No image) 

Social Network Expanded.png
Gephi's Network Graph

Zoomed in Directed Graph Our Simulation.png
Zoomed In

Sample Data


Breakout Level 2