Max-Planck-Institut für Informatik
max planck institut
mpii logo Minerva of the Max Planck Society

MPI-INF or MPI-SWS or Local Campus Event Calendar

<< Previous Entry Next Entry >> New Event Entry Edit this Entry Login to DB (to update, delete)
What and Who
Title:Learning-Based Synthesis
Speaker:Daniel Neider
coming from:UCLA
Speakers Bio:I work as postdoctoral researcher in the ExCAPE project at University of
Illinois at Urbana-Champaign and University of California, Los Angeles. I
joined ExCAPE in August 2014.
I received my Ph.D. from RWTH Aachen University in April 2014, where I
worked with Christof Löding and Wolfgang Thomas.
My thesis is on Applications of Automata Learning in Verification and
Synthesis. During this time, I visited Prof. Madhusudan at University of
Illinois at Urbana-Champaign.
I graduated from RWTH Aachen University with a Master of Science in
Computer Science in November 2007.
Event Type:SWS Colloquium
Visibility:D1, D2, D3, D4, D5, SWS, RG1, MMCI
We use this to send out email in the morning.
Level:AG Audience
Date, Time and Location
Date:Monday, 25 July 2016
Duration:60 Minutes
Synthesis, the automatic construction of objects related to hard- and
software, is one of the great challenges of computer science. Although
synthesis problems are impossible to solve in general, learning-based
approaches, in which the synthesis of an object is based on learning
from examples, have recently been used to build elegant and extremely
effective solutions for a large number of difficult problems. Such
examples include automatically fixing bugs, translating programs from
one language into another, program verification, as well as the
generation of high-level code from given specifications.

This talk gives an introduction to learning-based synthesis. First, we
develop a generic view on learning-based synthesis, called abstract
learning frameworks for synthesis, which introduces a common
terminology to compare and contrast learning-based synthesis
techniques found in the literature. Then, we present a learning-based
program verifier, which can prove the correctness of numeric programs
(nearly) automatically, and show how this technique can be modeled as
an abstract learning framework for synthesis. During the talk, we
present various examples that highlight the power of the
learning-based approach to synthesis.
Name(s):Vera Schreiber
EMail:--email address not disclosed on the web
Video Broadcast
Video Broadcast:YesTo Location:Saarbrücken
To Building:E1 5To Room:029
Tags, Category, Keywords and additional notes
Attachments, File(s):
Vera Schreiber/MPI-SWS, 07/19/2016 09:52 AM
Last modified:
Uwe Brahm/MPII/DE, 11/24/2016 04:13 PM
  • Vera Schreiber, 07/19/2016 10:03 AM -- Created document.