Optimization by Move–Class Deflation
نویسندگان
چکیده
A new approach to combinatorial optimization based on systematic move– class deflation is proposed. The algorithm combines heuristics of genetic algorithms and simulated annealing, and is mainly entropy–driven. It is tested on two problems known to be NP hard, namely the problem of finding ground states of the SK spin–glass and of the 3-D ±J spin–glass. The algorithm is sensitive to properties of phase spaces of complex systems other than those explored by simulated annealing, and it may therefore also be used as a diagnostic instrument. Moreover, dynamic freezing transitions, which are well known to hamper the performance of simulated annealing in the large system limit are not encountered by the present setup.
منابع مشابه
Deflation Methods for Sparse PCA
In analogy to the PCA setting, the sparse PCA problem is often solved by iteratively alternating between two subtasks: cardinality-constrained rank-one variance maximization and matrix deflation. While the former has received a great deal of attention in the literature, the latter is seldom analyzed and is typically borrowed without justification from the PCA context. In this work, we demonstra...
متن کاملBounded Anytime Deflation
This paper rearranges a key property of anytime algorithms to derive a technique that determines an admissible lower bound on the minimum solution cost. In conjunction with a class of performance metric known as ‘time-focussed metrics’, the technique can be used to determine when to abort the deflation process within an anytime planning algorithm. The aim of this paper is to exploit this techni...
متن کاملNewton's method with deflation for isolated singularities of polynomial systems
We present a modification of Newton’s method to restore quadratic convergence for isolated singular solutions of polynomial systems. Our method is symbolic-numeric: we produce a new polynomial system which has the original multiple solution as a regular root. We show that the number of deflation stages is bounded by the multiplicity of the isolated root. Our implementation performs well on a la...
متن کاملDeflating polygons to the limit
In this paper we study polygonal transformations through an operation called deflation. It is known that some families of polygons deflate infinitely for given deflation sequences. Here we show that every infinite deflation sequence of a polygon P has a unique limit, and that this limit is flat if and only if exactly two vertices of P move (are reflected) finitely many times in the sequence.
متن کاملGeneralized Deflation Algorithms for the Blind Source-factor Separation of Mimo-fir Channels
The present paper proposes a class of iterative deflation algorithms to solve the blind source-factor separation for the outputs of multiple-input multiple-output finite impulse response (MIMO-FIR) channels. Using one of the proposed deflation algorithms, filtered versions of the source signals, each of which is the contribution of each source signal to the outputs of MIMO-FIR channels, are ext...
متن کامل