TY - JOUR
T1 - Delayed information and action in on-line algorithms
AU - Albers, Susanne
AU - Charikar, Moses
AU - Mitzenmacher, Michael
N1 - Funding Information:
1A preliminary version of this paper was presented at the 39th Annual Symposium on Foundations of Computer Science (FOCS), 1998. 2Part of this work was done while at the Max-Planck-Institut für Informatik, Saarbrücken, Germany. 3Supported by a Stanford Graduate Fellowship, an ARO MURI Grant DAAH04-96-1-0007 and NSF Award CCR-9357849, with matching funds from IBM, Schlumberger Foundation, Shell Foundation, and Xerox Corporation. 4 Supported by NSF CAREER Grant CCR-9983832 and an Alfred P. Sloan Research Fellowship. Most of this work was done while employed at Compaq Systems Research Center.
PY - 1998
Y1 - 1998
N2 - Most on-line analysis assumes that, at each time-step, all relevant information up to that time step is available and a decision has an immediate effect. In many on-line problems, however, the time relevant information is available and the time a decision has an effect may be decoupled. For example, when making an investment, one might not have completely up-to-date information on market prices. Similarly, a buy or sell order might only be executed some time later in the future. We introduce and explore natural delayed models for several well-known on-line problems. Our analyses demonstrate the importance of considering timeliness in determining the competitive ratio of an on-line algorithm. For many problems, we demonstrate that there exist algorithms with small competitive ratios even when large delays affect the timeliness of information and the effect of decisions.
AB - Most on-line analysis assumes that, at each time-step, all relevant information up to that time step is available and a decision has an immediate effect. In many on-line problems, however, the time relevant information is available and the time a decision has an effect may be decoupled. For example, when making an investment, one might not have completely up-to-date information on market prices. Similarly, a buy or sell order might only be executed some time later in the future. We introduce and explore natural delayed models for several well-known on-line problems. Our analyses demonstrate the importance of considering timeliness in determining the competitive ratio of an on-line algorithm. For many problems, we demonstrate that there exist algorithms with small competitive ratios even when large delays affect the timeliness of information and the effect of decisions.
UR - http://www.scopus.com/inward/record.url?scp=0032306182&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:0032306182
SN - 0272-5428
SP - 71
EP - 80
JO - Annual Symposium on Foundations of Computer Science - Proceedings
JF - Annual Symposium on Foundations of Computer Science - Proceedings
T2 - Proceedings of the 1998 39th Annual Symposium on Foundations of Computer Science
Y2 - 8 November 1998 through 11 November 1998
ER -