TY - GEN
T1 - Structured hedging for resource allocations with leverage
AU - Johnson, Nicholas
AU - Banerjee, Arindam
N1 - Publisher Copyright:
© 2015 ACM.
Copyright:
Copyright 2016 Elsevier B.V., All rights reserved.
PY - 2015/8/10
Y1 - 2015/8/10
N2 - Data mining algorithms for computing solutions to online resource allocation (ORA) problems have focused on budgeting resources currently in possession, e.g., investing in the stock market with cash on hand or assigning current employees to projects. In several settings, one can leverage borrowed resources with which tasks can be accomplished more efficiently and cheaply. Additionally, a variety of opposing allocation types or positions may be available with which one can hedge the allocation to alleviate risk from external changes. In this paper, we present a formulation for hedging online resource allocations with leverage and propose an efficient data mining algorithm (SHERAL). We pose the problem as a constrained online convex optimization problem. The key novel components of our formulation are (1) a loss function for general leveraging and opposing allocation positions and (2) a penalty function which hedges between structurally dependent allocation positions to control risk. We instantiate the problem in the context of portfolio selection and evaluate the effectiveness of the formulation through extensive experiments on five datasets in comparison with existing algorithms and several variants.
AB - Data mining algorithms for computing solutions to online resource allocation (ORA) problems have focused on budgeting resources currently in possession, e.g., investing in the stock market with cash on hand or assigning current employees to projects. In several settings, one can leverage borrowed resources with which tasks can be accomplished more efficiently and cheaply. Additionally, a variety of opposing allocation types or positions may be available with which one can hedge the allocation to alleviate risk from external changes. In this paper, we present a formulation for hedging online resource allocations with leverage and propose an efficient data mining algorithm (SHERAL). We pose the problem as a constrained online convex optimization problem. The key novel components of our formulation are (1) a loss function for general leveraging and opposing allocation positions and (2) a penalty function which hedges between structurally dependent allocation positions to control risk. We instantiate the problem in the context of portfolio selection and evaluate the effectiveness of the formulation through extensive experiments on five datasets in comparison with existing algorithms and several variants.
KW - Finance
KW - Online learning
KW - Structured learning
UR - http://www.scopus.com/inward/record.url?scp=84954088744&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84954088744&partnerID=8YFLogxK
U2 - 10.1145/2783258.2783378
DO - 10.1145/2783258.2783378
M3 - Conference contribution
AN - SCOPUS:84954088744
T3 - Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
SP - 477
EP - 486
BT - KDD 2015 - Proceedings of the 21st ACM SIGKDD Conference on Knowledge Discovery and Data Mining
PB - Association for Computing Machinery
T2 - 21st ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD 2015
Y2 - 10 August 2015 through 13 August 2015
ER -