 |
Center for Nonlinear Studies |
> In this talk I will introduce data stream algorithms, and present the motivation, theory, and some of the applications behind them. The study of data stream algorithms largely began with the seminal paper of Alon, Matias, and Szegedy, and in the decade since then the area has received a lot of research attention in line with the explosive growth of information streams on the Internet. Examples of such streams include update streams in social networks, traffic flowing through backbone routers on the Internet, search queries, database updates etc. The core problem in data stream algorithms research is to estimate useful statistics of the stream in an online fashion. The challenge is to deal with the massive sizes of the streams, and perform the computation in a memory efficient manner. I will introduce the model for data stream computation, some of its variants, and present some of our results. I will also present some of the applications in the aforementioned examples. >
Host: Shiva Kasiviswanathan