Use of networks to improve prediction of mathematical models in the era of Big Data

Title:Use of networks to improve prediction of mathematical models in the era of Big Data

时 间：2015年6月11日（周四） 下午15:00-16:00

地 点：理学院二楼会议室

主讲人：王海燕，ArizonaState University

Abstract

Therich and big data generated and read by millions of users on social media tellswhat is happening in the real world in a rapid and accurate fashion. In recent years many researchers haveexplored real-time streaming data from Twitter for a broad range ofapplications such as predicting flu trends. In this talk, we present our design and implementation of a prototypesystem to collect flu related twitter data. Further we use partial differential equation models to describe thediffusion of flu related information in Twitter in both spatial and temporaldimensions. We correlate the resultswith official statistics from Center for Disease Control and Prevention (CDC). These results demonstrate that the system canbe used to real-timely monitor spread of flu trends.