Title :
Convex duality and the value of side information
Author_Institution :
Inf. Syst. Lab., Stanford Univ., CA, USA
fDate :
29 Jun-4 Jul 1997
Abstract :
Let X be a random variable with distribution P on a measurable space χ. A function q*(x) is log-optimum in a convex family 𝒬 of nonnegative measurable functions on χ if it attains the maximum growth exponent W𝒬(X)=supq(xin𝒬)E{log q(x)}. The analysis of random variable with joint distributions generalizes to the setting of stationary processes and log-optimum selections in multiplicative sequences of convex families
Keywords :
channel capacity; random processes; sequences; statistical analysis; convex duality; convex families; information rates; joint distributions; log-optimum selections; maximum growth exponent; measurable space; multiplicative sequences; nonnegative measurable functions; random variable; side information; stationary processes; Extraterrestrial measurements; Information systems; Investments; Kernel; Maximum likelihood estimation; Q measurement; Random variables; Rate-distortion; Weight measurement;
Conference_Titel :
Information Theory. 1997. Proceedings., 1997 IEEE International Symposium on
Conference_Location :
Ulm
Print_ISBN :
0-7803-3956-8
DOI :
10.1109/ISIT.1997.613198
