TY - GEN
T1 - On the α-loss Landscape in the Logistic Model
AU - Sypherd, Tyler
AU - Diaz, Mario
AU - Sankar, Lalitha
AU - Dasarathy, Gautam
N1 - Publisher Copyright:
© 2020 IEEE.
Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.
PY - 2020/6
Y1 - 2020/6
N2 - We analyze the optimization landscape of a recently introduced tunable class of loss functions called α-loss, α (0, ∞], in the logistic model. This family encapsulates the exponential loss (α = 1/2), the log-loss (α = 1), and the 0-1 loss (α = ∞) and contains compelling properties that enable the practitioner to discern among a host of operating conditions relevant to emerging learning methods. Specifically, we study the evolution of the optimization landscape of α-loss with respect to α using tools drawn from the study of strictly-locally-quasi-convex functions in addition to geometric techniques. We interpret these results in terms of optimization complexity via normalized gradient descent.
AB - We analyze the optimization landscape of a recently introduced tunable class of loss functions called α-loss, α (0, ∞], in the logistic model. This family encapsulates the exponential loss (α = 1/2), the log-loss (α = 1), and the 0-1 loss (α = ∞) and contains compelling properties that enable the practitioner to discern among a host of operating conditions relevant to emerging learning methods. Specifically, we study the evolution of the optimization landscape of α-loss with respect to α using tools drawn from the study of strictly-locally-quasi-convex functions in addition to geometric techniques. We interpret these results in terms of optimization complexity via normalized gradient descent.
UR - http://www.scopus.com/inward/record.url?scp=85090404551&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85090404551&partnerID=8YFLogxK
U2 - 10.1109/ISIT44484.2020.9174356
DO - 10.1109/ISIT44484.2020.9174356
M3 - Conference contribution
AN - SCOPUS:85090404551
T3 - IEEE International Symposium on Information Theory - Proceedings
SP - 2700
EP - 2705
BT - 2020 IEEE International Symposium on Information Theory, ISIT 2020 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2020 IEEE International Symposium on Information Theory, ISIT 2020
Y2 - 21 July 2020 through 26 July 2020
ER -