[DL] Open PhD Position
r.f.moeller at tu-harburg.de
Thu Feb 10 16:15:10 CET 2011
The research group on Logic-Based Computational Intelligence at Hamburg University of Technology offers a post in the project "PrIAS: Probabilistic Reasoning for Intelligent Autonomous Systems". Probabilistic extensions of predicate logic are used to model the abduction-based interpretation of input streams (or channels) of ground terms, which are attached with time stamps, and which represent, for instance, low level input from sensor networks (Cyber-Physical Systems scenario) or video surveillance systems (Computer Vision scenario). As a representation language used by an interpretation agent we investigate a combination of description logics and Horn logic. Probability theory is used to score different interpretation alternatives. Given distributions of single events in a sequence, a new variant of the Viterbi algorithm is to be developed to compute the most likely sequence of interpretations. Extending previous approaches we consider first-order structures in this context rather than propositional logic as usual. Based on interpretation results the agent computes optimal actions, e.g., in order to move in a cyber-physical system.
Candidates are expected to be able to provide teaching support for lectures with the following topics:
Computational Logic, Multimedia Information Extraction and Retrieval, Foundations of Machine Learning and Data Mining,
Architecture and Implementation of Database Systems, Introduction to Databases (in German), and Intelligent Autonomous Agents
The project starts 1. April 2011 and ends 31. March 2015. Salary will be in the range of 35-000 to 40.000 EUR
depending on age and qualification (EGr. 13 TV-L).
University regulations require candidates to hold a Diploma or Master Degree in Computer Science. We prefer candidates
with a strong focus on Computational Logics and Multi-Agent Theory.
Please send applications by email to
Prof. Dr. Ralf Möller
Hamburg University of Technology
Tel.: 040 42878 3462
Email: r.f.moeller at tu-harburg.de
More information about the dl