Performance and stability of many iterative algorithms such as stochastic gradient descent largely depend on a fixed and scalar step-size parameter. Use of a fixed and scalar step-size value may lead to limited performance in many problems. We study several existing step-size adaptation algorithms in nonstationary, supervised learning problems using simulated and real-world data. We discover that effectiveness of the existing step-size adaptation algorithms requires tuning of a meta parameter across problems. We introduce a new algorithm - Autostep - by combining several new techniques with an existing algorithm, and demonstrate that it can effectively adapt a vector step-size parameter on all of our training and test problems without tuning its meta parameter across them. Autostep is the first step-size adaptation algorithm that can be used in widely different problems with the same setting of all of its parameters.
Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:AEU.10048/1550 |
Date | 11 1900 |
Creators | Mahmood, Ashique |
Contributors | Sutton, Richard S. (Computing Science), Sutton, Richard S. (Computing Science), Shah, Sirish L. (Chemical and Materials Engineering), Schuurmans, Dale (Computing Science) |
Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
Language | en_US |
Detected Language | English |
Type | Thesis |
Format | 4528454 bytes, application/pdf |
Page generated in 0.0025 seconds