Generating exact nonlinear ranking functions by symbolic-numeric hybrid method

@article{Shen2013GeneratingEN,
  title={Generating exact nonlinear ranking functions by symbolic-numeric hybrid method},
  author={Li-Yong Shen and Min Wu and Zhengfeng Yang and Zhenbing Zeng},
  journal={Journal of Systems Science and Complexity},
  year={2013},
  volume={26},
  pages={291-301}
}
  • L. Shen, Min Wu, Zhenbing Zeng
  • Published 2 April 2013
  • Computer Science, Mathematics
  • Journal of Systems Science and Complexity
This paper presents a hybrid symbolic-numeric algorithm to compute ranking functions for establishing the termination of loop programs with polynomial guards and polynomial assignments. The authors first transform the problem into a parameterized polynomial optimization problem, and obtain a numerical ranking function using polynomial sum-of-squares relaxation via semidefinite programming (SDP). A rational vector recovery algorithm is deployed to recover a rational polynomial from the numerical… 

Proving total correctness and generating preconditions for loop programs via symbolic-numeric computation methods

We present a symbolic-numeric hybrid method, based on sum-of-squares (SOS) relaxation and rational vector recovery, to compute inequality invariants and ranking functions for proving total

SVMRanker: a general termination analysis framework of loop programs via SVM

TLDR
SVMRanker is presented, a general framework for proving termination of programs, which is able to synthesize different types of ranking functions for programs with both linear and polynomial updates, based on Support-Vector Machines (SVM).

Synthesizing Nested Ranking Functions for Loop Programs via SVM

TLDR
This paper considers nested ranking functions for loop programs and shows that the existence problem of a nested ranking function is equivalent to the existenceProblem of a hyperplane separating classes of data, which allows us to leverage Support-Vector Machines techniques for the synthesis of nestedranking functions.

Non-polynomial Worst-Case Analysis of Recursive Programs

TLDR
This work applies ranking functions to recursion, resulting in measure functions, and can synthesize bounds of various forms including O(n log n) and O(nr), where r is not an integer.

Ranking Function Detection via SVM: A More General Method

TLDR
This paper proposes an approach to the synthesis of ranking functions for loops via support vector machine (SVM), and transforms the ranking function detection problem into a binary classification problem.

Termination Analysis of Probabilistic Programs Through Positivstellensatz's

TLDR
The approach synthesizes polynomial ranking-supermartingales through Positivstellensatz’s, yielding an efficient method which is not only sound, but also semi-complete over a large subclass of programs.

Synthesis of ranking functions via DNN

TLDR
This paper proposes a new approach to synthesis of non-polynomial ranking functions for loops via deep neural network (DNN), and builds a ranking function template by DNN structure which can be learned by the train-set and verified by a new verification method.

Automated Recurrence Analysis for Almost-Linear Expected-Runtime Bounds

TLDR
The approach can infer the asymptotically optimal expected-runtime bounds for recurrences of classical randomized algorithms, including RANDOMIZED-SEARCH, QUICK-SORT,QUICK-SELECT, COUPONCOLLECTOR, where the worst-case bounds are either inefficient, quadratic as compared to linear or almost-linear, or ineffective.

Efficient Algorithms for Checking Fast Termination in VASS

TLDR
It is shown that singularities in the normal are the key reason for asymptotic bounds such as exponential and non-elementary for VASS, and that the asymPTotic complexity bound is always polynomial and of the form ${\Theta}(n^k)$, for some k $\leq$ d.

Average-case analysis of programs: Automated recurrence analysis for almost-linear bounds

  • Computer Science
  • 2020
TLDR
The approach can infer the asymptotically optimal average-case bounds for classical randomized algorithms, including RANDOMIZED-SEARCH, QUICKSORT,QUICK-SELECT, COUPON-COLLECTOR, where the worstcase bounds are either inefficient, quadratic as compared to linear or almost-linear of average- case, or ineffective.

References

SHOWING 1-10 OF 19 REFERENCES

Practical Methods for Proving Program Termination

We present two algorithms to prove termination of programs by synthesizing linear ranking functions. The first uses an invariant generator based on iterative forward propagation with widening and

A Complete Method for the Synthesis of Linear Ranking Functions

TLDR
An automated method for proving the termination of an unnested program loop by synthesizing linear ranking functions is presented and it is shown that if a linear ranking function exists then it will be discovered by the method.

Synthesis of Linear Ranking Functions

TLDR
An algorithm is presented to synthesize linear ranking functions that can establish termination of program cycles and the representation of systems of linear inequalities and sets of linear expressions as polyhedral cones allows this search to be reduced to the computation of polars, intersections and projections ofpolyhedral cones.

Discovering Non-linear Ranking Functions by Solving Semi-algebraic Systems

Differing from [6] this paper reduces non-linear ranking function discovering for polynomial programs to semi-algebraic system solving, and demonstrates how to apply the symbolic computation tools,

DISCOVERER: a tool for solving semi-algebraic systems

  • Bican Xia
  • Mathematics, Computer Science
    ACCA
  • 2007
TLDR
The aim of this package is showing how an unified tolerance mechanism that was introduced for the SNAP package works, and with that, approximate calculations under certified tolerances without special skills in symbolic-numeric algebra can be continued.

Proving Program Invariance and Termination by Parametric Abstraction, Lagrangian Relaxation and Semidefinite Programming

TLDR
This new approach exploits the recent progress in the numerical resolution of linear or bilinear matrix inequalities by semidefinite programming using efficient polynomial primal/dual interior point methods generalizing those well-known in linear programming to convex optimization.

Termination Analysis of Integer Linear Loops

TLDR
Usually, ranking function synthesis and invariant generation over a loop with integer variables involves abstracting the loop to have real variables, but this work introduces a technique for handling loops over integer variables directly.

The computational complexity of simultaneous Diophantine approximation problems

  • J. Lagarias
  • Mathematics, Computer Science
    23rd Annual Symposium on Foundations of Computer Science (sfcs 1982)
  • 1982
TLDR
It is shown that the problem of deciding whether a given vector α of rational numbers has a simultaneous approximation of specified accuracy with respect to the sup norm with denominator Q in a given interval 1 ≤ Q ≤ N is NP-complete.

Termination of Integer Linear Programs

TLDR
It is shown that termination of deterministic linear loops is decidable over the integers in the homogeneous case, and over the rationals in the general case, by analyzing the powers of a matrix symbolically using its eigenvalues.