conclusions

Document Sample
conclusions Powered By Docstoc
					                                    Conclusions

• overview and survey from a logical / ILP
  perspective
• Distinction between




                                                    ICML-Tutorial, Banff, Canada, 2004
  – Model-based: BN, PLPs, PRMs,BLPs,...
  – vs. proof-based: SCFGs, SLPs, PRISM, ...
• Learning Settings:
  – Learning from interpretations: PLPs,PRMs,BLPs
  – Learning from entailment: SLPs, PRISM
  – Learning from traces: RMMs, LOHMMs
                        Conclusions - continued

• Learning includes principles from
  – Inductive logic learning / multi-relational data
    mining




                                                       ICML-Tutorial, Banff, Canada, 2004
     • Refinement operators
     • Background knowledge
     • Bias
  – Statistical learning
     • Likelihood
     • Independencies
     • Priors
                       Thank you for your ...




                                                   ICML-Tutorial, Banff, Canada, 2004
Sorry for all the probabilistic, logical stuff !
We hope that you have learned something !
                                                    Selected Links
• Conferences, Workshops & Summer Schools
   – AAAI-2000 workshop on "Learning Statistical Models from Relational Data"
     (SRL-2000)
   – Summer School on Relational Data Mining 2002
   – IJCAI-2003 workshop on "Learning Statistical Models from Relational Data"
     (SRL-2003)




                                                                                     ICML-Tutorial, Banff, Canada, 2004
   – ICML-2004 workshop on "Statistical Relational Learning and its Connections to
     Other Fields" (SRL-2004)
   – Forthcoming Dagstuhl seminar on
     "Probabilistic, Logical and Relational Learning - Towards a Synthesis"
• Systems & Data
   – Probabilistic-Logical Model Repository
• Projects
   – Evidence Extraction and Link Discouvery (EELD) DARPA Program
   – Efficient first-order probabilistic models for inference and learning, EPSRC
     research grant GR/N0739
   – Application of Probabilistic Inductive Logic Programming (APRIL I) European
     Union Assessment Project IST-2001-33035
   – Application of Probabilistic Inductive Logic Programming (APRIL II) Specific
     Targeted Research Project" funded by the European Commission under the
     "Sixth Framework Programme (2002-2006); Information Society Technologies"
     "Future and Emerging Technologies" arm. Contract no. FP6-508861
                                     Selected Publications




ICML-Tutorial, Banff, Canada, 2004
                                     Selected Publications




ICML-Tutorial, Banff, Canada, 2004
                                     Selected Publications




ICML-Tutorial, Banff, Canada, 2004
                                     Selected Publications




ICML-Tutorial, Banff, Canada, 2004

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:4
posted:9/14/2012
language:English
pages:8