Content Gazette - October 2015 - 25

Call	
  for	
  Papers
IEEE	
  Signal	
  Processing	
  Society
IEEE	
  Transactions	
  on	
  Signal	
  and	
  Information	
  Processing	
  over	
  Networks
	
  
SPECIAL	
  ISSUE	
  ON	
  INFERENCE	
  AND	
  LEARNING	
  OVER	
  NETWORKS	
  

	
  
Networks	
   are	
   everywhere.	
   They	
   surround	
   us	
   at	
   different	
   levels	
   and	
   scales,	
   whether	
   we	
   are	
   dealing	
   with	
   communications	
  
networks,	
  power	
  grids,	
  biological	
  colonies,	
  social	
  networks,	
  sensor	
  networks,	
  or	
  distributed	
  Big	
  Data	
  depositories.	
  Therefore,	
  
it	
  is	
  not	
  hard	
  to	
  appreciate	
  the	
  ongoing	
  and	
  steady	
  progression	
  of	
  network	
  science,	
  a	
  prolific	
  research	
  field	
  spreading	
  across	
  
many	
  theoretical	
  as	
  well	
  as	
  applicative	
   domains.	
  Regardless	
  of	
  the	
  particular	
  context,	
  the	
  very	
  essence	
  of	
  a	
  network	
  resides	
  in	
  
the	
   interaction	
   among	
   its	
   individual	
   constituents,	
   and	
   Nature	
   itself	
   offers	
   beautiful	
   paradigms	
   thereof.	
   Many	
   biological	
  
networks	
   and	
   animal	
   groups	
   owe	
   their	
   sophistication	
   to	
   fairly	
   structured	
   patterns	
   of	
   cooperation,	
   which	
   are	
   vital	
   to	
   their	
  
successful	
   operation.	
   While	
   each	
   individual	
   agent	
  is	
   not	
   capable	
   of	
  sophisticated	
  behavior	
   on	
   its	
   own,	
  the	
   combined	
   interplay	
  
among	
   simpler	
  units	
  and	
   the	
  distributed	
  processing	
  of	
  dispersed	
  pieces	
  of	
  information,	
  enable	
  the	
  agents	
   to	
  solve	
  complex	
  
tasks	
   and	
   enhance	
   dramatically	
   their	
   performance.	
   Self--‐organization,	
   cooperation	
   and	
   adaptation	
   emerge	
   as	
   the	
   essential,	
  
combined	
   attributes	
   of	
   a	
   network	
   tasked	
   with	
   distributed	
   information	
   processing,	
   optimization,	
   and	
   inference.	
   Such	
   a	
  
network	
  is	
  conveniently	
  described	
  as	
  an	
  ensemble	
  of	
  spatially	
  dispersed	
  (possibly	
  moving)	
  agents,	
  linked	
  together	
  through	
  a	
  
(possibly	
  time--‐varying)	
   connection	
   topology.	
  The	
   agents	
   are	
   allowed	
  to	
  interact	
  locally	
  and	
  to	
  perform	
  in--‐network	
  processing,	
  
in	
   order	
   to	
   accomplish	
   the	
   assigned	
   inferential	
   task.	
   Correspondingly,	
   several	
   problems	
   such	
   as,	
   e.g.,	
   network	
   intrusion,	
  
community	
  detection,	
  and	
  disease	
  outbreak	
  inference,	
  can	
  be	
   conveniently	
  described	
  by	
  signals	
  on	
  graphs,	
  where	
  the	
  graph	
  
typically	
   accounts	
   for	
   the	
   topology	
   of	
   the	
   underlying	
   space	
   and	
   we	
   obtain	
   multivariate	
   observations	
   associated	
   with	
  
nodes/edges	
   of	
   the	
   graph.	
   The	
   goal	
   in	
   these	
   problems	
   is	
   to	
   identify/infer/learn	
   patterns	
   of	
   interest,	
   including	
   anomalies,	
  
outliers,	
   and	
   existence	
   of	
   latent	
   communities.	
   Unveiling	
   the	
   fundamental	
   principles	
   that	
   govern	
   distributed	
   inference	
   and	
  
learning	
   over	
   networks	
   has	
   been	
   the	
   common	
   scope	
   across	
   a	
   variety	
   of	
   disciplines,	
   such	
   as	
   signal	
   processing,	
   machine	
  
learning,	
   optimization,	
   control,	
   statistics,	
   physics,	
   economics,	
   biology,	
  computer,	
   and	
   social	
   sciences.	
   In	
   the	
   realm	
   of	
   signal	
  
processing,	
   many	
   new	
   challenges	
   have	
   emerged,	
   which	
   stimulate	
   research	
   efforts	
   toward	
   delivering	
   the	
   theories	
   and	
  
algorithms	
   necessary	
   to	
   (a)	
   designing	
   networks	
   with	
   sophisticated	
   inferential	
   and	
   learning	
   abilities;	
   (b)	
   promoting	
   truly	
  
distributed	
   implementations,	
   endowed	
   with	
   real--‐time	
   adaptation	
   abilities,	
   needed	
   to	
   face	
   the	
   dynamical	
   scenarios	
   wherein	
  
real--‐world	
  networks	
  operate;	
  and	
  (c)	
  discovering	
  and	
  disclosing	
  significant	
  relationships	
  possibly	
  hidden	
  in	
  the	
  data	
  collected	
  
from	
   across	
   networked	
   systems	
   and	
   entities.	
   This	
   call	
   for	
   papers	
   therefore	
   encourages	
   submissions	
   from	
   a	
   broad	
   range	
   of	
  
experts	
  that	
  study	
  such	
  fundamental	
  questions,	
  including	
  but	
  not	
  limited	
  to:	
  
*
*
*
*
*
*
*
*
*
*
*
*

Adaptation	
  and	
  learning	
  over	
  networks.	
  	
  
Consensus	
  strategies;	
  diffusion	
  strategies.	
  
Distributed	
  detection,	
  estimation	
  and	
  filtering	
  over	
  networks.	
  	
  
Distributed	
  dictionary	
  learning.	
  
Distributed	
  game--‐theoretic	
  learning.	
  
Distributed	
  machine	
  learning;	
  online	
  learning.	
  	
  
Distributed	
  optimization;	
  stochastic	
  approximation.	
  	
  
Distributed	
  proximal	
  techniques,	
  sub--‐gradient	
  techniques.	
  
Learning	
  over	
  graphs;	
  network	
  tomography.	
  
Multi--‐agent	
  coordination	
  and	
  processing	
  over	
  networks.	
  	
  
Signal	
  processing	
  for	
  biological,	
  economic,	
  and	
  social	
  networks.	
  	
  
Signal	
  processing	
  over	
  graphs.	
  

Prospective	
   authors	
   should	
   visit	
   http://www.signalprocessingsociety.org/publications/periodicals/tsipn/	
   for	
   information	
   on	
  
paper	
  submission.	
  Manuscripts	
  should	
  be	
  submitted	
  via	
  Manuscript	
  Central	
  at	
   http://mc.manuscriptcentral.com/tsipn--‐ieee.	
  
	
  
Important	
  Dates:	
  
* Manuscript	
  submission:	
  February	
  1,	
  2016	
  
* First	
  review	
  completed:	
  April	
  1,	
  2016	
  
* Revised	
  manuscript	
  due:	
  May	
  15,	
  2016	
  
* Second	
  review	
  completed:	
  July	
  15,	
  2016	
  
* Final	
  manuscript	
  due:	
  September	
  15,	
  2016	
  
* Publication:	
  December	
  1,	
  2016	
  
	
  
Guest	
  Editors:	
  
	
   Vincenzo	
  Matta,	
  University	
  of	
  Salerno,	
  Italy,	
  vmatta@unisa.it	
  	
  
	
   Cédric	
  Richard,	
  University	
  of	
  Nice	
  Sophia--‐Antipolis,	
  France,	
  cedric.richard@unice.fr	
  	
  	
  
	
   Venkatesh	
  Saligrama,	
  Boston	
  University,	
  USA,	
  srv@bu.edu	
  	
  	
  
	
   Ali	
  H.	
  Sayed,	
  University	
  of	
  California,	
  Los	
  Angeles,	
  USA,	
  sayed@ucla.edu	
  	
  

www.signalprocessingsociety.org

[25]

OCTOBER 2015
http://www.signalprocessingsociety.org/publications/periodicals/tsipn/ http://mc.manuscriptcentral.com/tsipn--‐ieee http://www.signalprocessingsociety.org
Table of Contents for the Digital Edition of Content Gazette - October 2015
