Biography
About Me
I'm a Research Assistant at College of Computer & Information Science, Northeastern University under the supervision of Professor David Choffnes.
Before joining Northeastern in the Fall 2015, I attended Lahore University of Management Sciences (LUMS) in Lahore, Pakistan and earned a Bachelor’s degree in Computer Science.
I'm passionate about design and implementation of tools that improve the privacy, reliability, and scalability of real world systems and networks by applying techniques used in Machine Learning and Big Data Analytics.
Experience
Current Research Projects
Passport: Country-level Router Geolocation
Internet routing is based on Internet Service Provider (ISP) policies. These policies are usually kept private by the ISPs. Current geolocation services use limited information and measurements to geolocate routers. Therefore, these current geolocation services have a low accuracy for router locations. In this project, we are using the applications of Machine Learning on existing, inaccurate geolocation sources, router information, and Internet measurements to develop an algorithm (and a system based on this algorithm) to address the task of predicting router locations. Utilizing this information, this project aims at quantifying the impact of foreign governments on Internet paths and developing a prototype to avoid them.
Source Code:
The source code and the data for our system will soon be public. Stay tuned!
Source Code Documentation:
The Developer's Guide of
REST API:
To access our public API for
Website:
Our website is now live in pre-alpha phase! The core functionality has been developed in Python (Flask, Scikit-Learn, and Django) and the Frontend was built using D3 and Bootstrap. We're adding more functionality by every passing day. Please visit our website at passport.ccs.neu.edu or click the image below.
Past Research Projects
Cyclone: Dynamic Cloud Virtualization
Cyclone is a dynamic cloud virtualization system that can manipulate the the cloud environment. It supports dynamic capabilities such as controlling time, shape and inputs. In this project, we aimed at leveraging large-scale system resources, program analysis and partial execution to ensure cloud replay in a cost-effective manner. It was a collaborative project between University of Illinois Urbana-Champaign (UIUC) and LUMS and a prototype implementation is still under works by the research groups.
Distributed and Secure Veriflow
Veriflow is an online networks invariant verifier developed by UIUC. It sits in as layer between the control and data plane to verify network invariants. Our research primarily focused on extending Veriflow to allow for a private and verifiable inter-domain routing (between different Veriflow instances) without compromising the space and time complexities of the verifying global network invariants (such as, checking for forwading loops and network black holes).
Disease Surveillance using Twitter
Twitter streams are used in various analyses, including and not limited to predicting election results and disease trends. Our work focused on the reliability of twitter as a source of disease surveillance. We gathered 2 million tweets for dengue and influenza during the year 2012-13, classified them in various groups and analyzed the usefulness of twitter as source of epidemic prediction, resource information, personal experience and advertisement in various countries for a given timeline. However, with the large amounts of noise in the data didn't, the results didn’t show a significant promise to target a publication.
Education
M.S. in Computer Science
College of Computer & Information Science, Northeastern University
B.S. in Computer Science
School of Sciences and Engineering, LUMS
Contact Info
Office:
Northeastern University
College of Computer & Information Science
460 ISEC
805 Columbus Ave
Boston, MA, 02120
E-mail:
muzammil *at* ccs [dot] neu [dot] edu