Cinque Terre

Sahar Mehrpour

Room 4402
Engineering Building, George Mason University
Fairfax, VA
email: smehrpour AT gmu DOT edu

About Me

I am a CS Ph.D. student at the George Mason University. I am working with Dr. Thomas LaToza in Developer Experience Design Lab. My research is designing and building tools for expediting program development.

Background

I received my bachelor's degree in Computer Science at Sharif University of Technology, Iran.

I continued my graduate studies at Sharif University of Technology under supervision of Dr. Alireza Zarei and at the University of Manitoba under supervision of Dr. Stephane Durocher. I was a member of the Computational Geometry Lab.

I was also a PhD student at the University of Utah before transferring to GMU.

Previous Projects

Clinical Data Visualization

This is a project is advised by Alexander Lex at the Scientific Computing and Imaging Institute at the University of Utah. I had a privilege to work on this project for one semester (Spring 2017). In this IRB-approved project, we tried to present a visualization of clinical data of patients. The target audiences for this visualization are medical doctors.

Custom Framework Phovea, TypeScript, Python, D3js, scikit-learn Python packages, CSS.

Uncertain Visualization

Collaborators: Navid Farhoudi

This project was done as a course project for Topological Data Analysis course at the University of Utah (Spring 2016). In this project, we present a visualization of ensemble vector fields taken from International Research Institute for Climate Society. In this project, we displayed flow, vector magnitude and Merge Tree of each ensemble.

JavaScript, D3js HTML, CSS

GitHub, Demo

Topic Modeling

Collaborators: Tamara Denning, Aniqua Baset

The conferences on Security and Privacy contain a lot of information. In this project we tried to visualize the topic modeling data in 4 different conferences in 30 years. The data is prepared by Aniqua Baset at the University of Utah (Fall 2016).

JavaScript, D3js, HTML, CSS

GitHub, Demo

Android Malware Detection

This project is done as a course project for Machine Learning course at the University of Utah (Fall 2016). The data in this project is labeled Android Applications and we wanted to detect Malware applications using classifiers. We implemented the well-known classifiers from scratch: ID3 (Decision Tree), Stochastic Gradient Descent For Least Squares, Perceptron, Balanced Winnow, SVM (Support Vector Machine), Linear Regression, Random Forest, AdaBoost.

Python

Dataset

Movie Miner

Collaborators: Sunipa Dev, Zahra Fahimfar, Siddartha Ravichandran

This project is done as a course project for a Visualization course at the University of Utah (Fall 2016). In this project, we present a visualization for a relatively big movie dataset. This dataset consists of more than 5 thousand movies. Each entry has around 20 attributes including Country, Director, Actor, Year, Genre, Rating, etc. MovieMiner is an interactive visualization for helping users filter and select desirable movies. This dataset is based on IMDB website.

JavaScript, D3js, HTML, CSS

GitHub, Demo, Video

KiDGraD

Collaborators: Yeganeh Bahoo, Stephane Durocher, Andrea Bunt

In this project we used Microsoft Kinect™ for Graph Drawing. We developed an application in which users can draw 2D graphs using body gestures. This project was a course project for HCI course at the University of Manitoba. We implemented the application in C# and used the information of skeleton detection for graph drawing.

C#, Microsoft Kinect™

Abstract, Poster, GitHub