Raja Velu graduated from University of Wisconsin-Madison in 1983; he taught in the UW system since then before moving to Syracuse University in 1998. He has been a visiting professor at Stanford University (2005-2016) in the statistics department. He has developed a course on Algorithmic Trading and Quantitative Strategies at Stanford. Raja is also affiliated with leading technology and financial companies; he served as the Forecasting lead at Yahoo! in the sponsored search and competitive intelligence areas. At Microsoft Research, he collaborated with researchers in the Search Labs to develop a Forecasting and recommendation system for high-end products called Prodcast. He also worked at IBM-Almaden and Google at Silicon Valley in the Data Science teams. Raja served as JPMC Faculty Fellow and spent a year with JPMC’s Electronic Client Solutions group in NYC.
His current research interests include large dimensional forecasting and big data issues in data science and conditional asset pricing and algorithmic trading in finance. His papers have appeared in leading journals such as Biometrika, Journal of Econometrics, Journal of Financial and Quantitative Analysis, Journal of the Royal Statistical Society etc. His book with Greg Reinsel and Kun Chen,Multivariate Reduced-Rank Regression: Theory and Applications published by Springer-Verlag is highly cited. His recent book, “Algorithmic Trading and Quantitative Strategies” with Maxence Hardy at JPMC and Daniel Nehren at ADIA,is published by Taylor and Francis. He has served as the Program Chair for Business and Economic Statistics section of ASA in 2003 JSM. Raja was recognized by Syracuse University with a Chancellor’s citation for his contribution to teaching, research and service in 2013.
Syracuse University, Syracuse, NY
February 23, 2024
Parsimonious Modeling and Forecasting of Chronological Data
An important feature of chronological data is the dependence of the present on the past. Extracting this dependence is very important to make reasonable predictions into the future. With the advent of electronic media, the data comes in high frequency and in high volume, transmitted instantaneously. In this talk, I will illustrate some applications with relevant methodologies. The high dimensionality is both a blessing and a curse. I will draw upon my research in the area of multivariate vector auto-regression and discuss canonical methods that both simplify the structure and provide better forecasts. The talk will cover practical issues involved in developing an agile forecasting system and some areas for future research.
