Priya Sangarneswaran

REU Researcher: Priya Sangarneswaran
Email:
Mentor: Peter Hammer, RUTCOR, Rutgers.
Email: hammer@rutcor.rutgers.edu

Priya worked with Drs. Peter Hammer and Ilya Muchnik, both of the Rutgers Center for Operations Research, as part of a team working on a long-term project in prediction based on performance patterns. To quote from her research report: "The work I did with RUTCOR this summer was not individual work, but work as part of a team. This was a great experience, not only because I learned a lot from the professors of RUTCOR, but also because it allowed me to be part of a project that had already begun years before I came here, and thus made me part of a team that was constantly looking for better results, new ways of analyzing the results, clear goals and purposes for their projects and challenging new studies that would be useful in answering real world questions. Working here at RUTCOR has definitely made me interested in pursuing a career in Operations Research."

The method of Logical Analysis of Data (LAD) provides techniques to identify various regularities in data; in particular, to predict some binary attribute on the basis of other attributes that may be continuous. Whereas all previous applications of LAD have been to reveal a cause-effect relationship, the objective of this work was to attempt to predict the future state of an attribute given the history of the states of a set of causal attributes. For this purpose, LAD has been applied to the Chinese Economy database -- the database on the marketization process in the industrial sector of the People's Republic of China. Twelve attributes had been observed in 29 provinces of China over 10 years from 1985 to 1994. Using this data they analyzed the ability of LAD to (1) predict labor productivity and (2) measure the importance of, and monotonicity trends of attributes in the prediction of labor productivity. Priya and Dr. Hammer are preparing a paper describing this research.


Index
WebMaster: dimacs-www@dimacs.rutgers.edu
Last modified January 21, 1997.