Network Analysis of US Critical Care Hospital Transfers
Course Project: 1.Developed an effective stochastic model to minimize the damage caused by the diffusion of virus in the network.
2. Developed a visualiztion tool which shows the spread of virus and the damage caused by it in the network.
Research: Adamic, L., Karkada, U.H., Iwashyna, T.J .The Potential Spread of Highly Resistant Hospital
Acquired Infections Via of Inter-Hospital Transfers of Critically Ill Patients.
Portfolio Analysis on finance.yahoo
This project presents an analysis of a portfolio of stocks with an explicit intent to guide your investments. Decision of what stocks to invest in is as much a function of changing economic conditions as much as your personal risk tolerance. Your risk tolerance may dependent on a number of personal or financial goals, and there often exists a dilemma whether to invest in a market with low-risk or high-risk.
Interactive Learning System (Development)
The Interactive Learning System makes an effort to expand education opportunities through online and distance learning experiences for teachers and students. (includes Text, File, Voice & Video transferring agent)
Google App Chat Application (Development)
A chat application developed using Google App Engine.
Recommender System Design for Social Networks
Abstract: Recommendation systems for Social networks are different from other kinds of systems, since the item recommended here are rational human beings rather than goods. Thus, the Social factor has to be accounted for when making a recommendation. In this paper, I have used a hybrid of Content and Collaborative filtering recommender algorithm to recommend new friends to a user in a Social network.
Reputation System Design for Decentralized P2P Networks
Abstract: In a decentralized P2P network there is no central body which regulates the actions and stores the reputation of the users. Trust plays a very important role when one is looking to download good content. The reputation system design proposed in this paper uses the social network property to determine the trust factor and calculate a relative reputation of every user in the network to a particular user.