DocumentCode
3632768
Title
Termination Analysis by Program Inversion
Author
Stefan Andrei
Author_Institution
Dept. of Comput. Sci., Lamar Univ., Beaumont, TX, USA
fYear
2008
Firstpage
277
Lastpage
280
Abstract
Automatic methods for discovering program runtime and proving program termination have always been a challenging problem in computer science. We present here a novel and systematic approach for calculating an upper bound of the maximum runtime of functions for a non-trivial class of programs. The proof is based on an induction over a treeof execution traces - a new mathematical data structure. As a consequence, this can also show termination for these functions. The approach uses symbolic-numeric algorithms over a novel mathematical data structure, and can systematically find the maximum runtime for a wide class of functions.
Keywords
"Runtime","Computer science","Scientific computing","Algorithm design and analysis","Upper bound","Tree data structures","Data structures","Partitioning algorithms","Arithmetic","Delay effects"
Publisher
ieee
Conference_Titel
Symbolic and Numeric Algorithms for Scientific Computing, 2008. SYNASC ´08. 10th International Symposium on
Print_ISBN
978-0-7695-3523-4
Type
conf
DOI
10.1109/SYNASC.2008.22
Filename
5204824
Link To Document