Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
Monotonic Networks Joseph Sill
 

Summary: Monotonic Networks
Joseph Sill
Computation and Neural Systems program
California Institute of Technology
MC 136­93, Pasadena, CA 91125
email: joe@cs.caltech.edu
Abstract
Monotonicity is a constraint which arises in many application do­
mains. We present a machine learning model, the monotonic net­
work, for which monotonicity can be enforced exactly, i.e., by virtue
of functional form. A straightforward method for implementingand
training a monotonic network is described. Monotonic networks
are proven to be universal approximators of continuous, differen­
tiable monotonic functions. We apply monotonic networks to a
real­world task in corporate bond rating prediction and compare
them to other approaches.
1 Introduction
Several recent papers in machine learning have emphasized the importance of pri­
ors and domain­specific knowledge. In their well­known presentation of the bias­
variance tradeoff (Geman and Bienenstock, 1992), Geman and Bienenstock conclude

  

Source: Abu-Mostafa, Yaser S. - Department of Mechanical Engineering & Computer Science Department, California Institute of Technology

 

Collections: Computer Technologies and Information Sciences