TR on connection pruning available

[Lutz Prechelt]

FTP-host: ftp.icsi.berkeley.edu
FTP-file: /pub/techreports/1995/tr-95-009.ps.Z
URL:      ftp://ftp.icsi.berkeley.edu/pub/techreports/1995/tr-95-009.ps.Z

The technical report 

   "Adaptive Parameter Pruning in Neural Networks"

is now available for anonymous ftp
from  ftp.icsi.berkeley.edu  in directory  /pub/techreports/1995/
as file  tr-95-009.ps.Z  (92 kB, 14 pages).


Here is the bibtex entry and abstract:

@TechReport{Prechelt95e,
  author =       "Lutz Prechelt",
  title =        "Adaptive Parameter Pruning in Neural Networks",
  institution =  "International Computer Science Institute",
  year =         1995,
  number =       "95-009",
  address =      "Berkeley, CA",
  month =        mar,
  Class =        "nn, learning, experiment, algorithm",
  URL = "ftp://ftp.icsi.berkeley.edu/pub/techreports/1995/tr-95-009.ps.Z},
  abstract = "Neural network pruning methods on the level of individual 
    network parameters (e.g. connection weights) can improve
    generalization. An open problem in the pruning methods
    known today (OBD, OBS, autoprune, epsiprune) is the selection of the
    number of parameters to be removed in each pruning step (pruning
    strength). This paper presents a pruning method \Def{lprune} that
    automatically adapts the pruning strength to the evolution of weights
    and loss of generalization
    during training. The method requires no algorithm parameter adjustment by
    the user.  The results of extensive experimentation indicate that lprune
    is often superior to autoprune (which is superior to OBD) on diagnosis
    tasks unless severe pruning early in the training process is required.
    Results of statistical significance tests comparing autoprune to the
    new method lprune as well as to backpropagation with early stopping
    are given for 14 different problems."
}

The ICSI internet connection is sometimes extremely slow and fails
often. If you have problems getting the document, just try again at
a different time.

Sorry, no hardcopies available from me.

  Lutz

Lutz Prechelt   (http://wwwipd.ira.uka.de/~prechelt/)   | Whenever you 
Institut fuer Programmstrukturen und Datenorganisation  | complicate things,
Universitaet Karlsruhe;  76128 Karlsruhe;  Germany      | they get
(Voice: +49/721/608-4068, FAX: +49/721/694092)          | less simple.