By Yuri Malitsky
This publication offers a modular and expandable approach within the quickly rising learn sector of computerized configuration and choice of the simplest set of rules for the example to hand. the writer provides the fundamental version in the back of ISAC after which info a couple of adjustments and functional functions. specifically, he addresses automatic function iteration, offline set of rules configuration for portfolio iteration, set of rules choice, adaptive solvers, on-line tuning, and parallelization.
The author's similar thesis was once honorably pointed out (runner-up) for the ACP Dissertation Award in 2014, and this publication contains a few accelerated sections and notes on fresh advancements. also, the thoughts defined during this booklet were effectively utilized to a couple of solvers competing within the SAT and MaxSAT overseas Competitions, profitable a complete of 18 gold medals among 2011 and 2014.
The ebook should be of curiosity to researchers and practitioners in synthetic intelligence, particularly within the region of laptop studying and constraint programming.
Read Online or Download Instance-Specific Algorithm Configuration PDF
Similar nonfiction_12 books
The Mysterious Island tells the intriguing tale of 5 american citizens stranded on an uncharted island within the South Pacific. in the course of the American Civil struggle, Richmond, Virginia was once the capital of the accomplice States of the US. 5 northern prisoners of conflict choose to break out Richmond in a slightly strange manner - through hijacking a balloon.
- Mechanisms, Transmissions and Applications: Proceedings of the Third MeTrApp Conference 2015
- Accelerated C♯ 2005
- Microchip Capillary Electrophoresis Protocols
- Hox Genes: Methods and Protocols
Extra info for Instance-Specific Algorithm Configuration
2 Minimizing a one-dimensional convex function by golden section section (see Fig. 2): two points X < Y are considered within the interval Œ0; 1 and their performance is measured as “pX ” and “pY ”. 1 X /m’ to 22 3 Instance-Specific Algorithm Configuration variable “b”. Now, if the function is indeed convex, if pX < pY (pX pY ), then the minimum of this one-dimensional function lies in the interval Œ0; Y (ŒX; 1). The search continues splitting the remaining interval (which shrinks geometrically fast) until the interval size “length” falls below a given threshold “ ”.
3S was the first sequential portfolio that won gold in more 46 5 ISAC for Algorithm Selection than one main category (SAT+UNSAT instances). The specifics of this solver are presented in Chap. 6. Although not explicitly shown, all the results are significant as per the Wilcoxon signed rank test with continuity correction. MSC is faster than SATzilla_R with p Ä 0:1 %. 3 Improved Algorithm Selection When compared with PSP, MSC replaced some of the default solvers for some clusters with other default solvers and, while lowering the training performance, this resulted in an improved test performance.
For each consequent step, the neighborhood is composed of all states obtained by adding or removing one subset from the current solution. The fitness function is then evaluated as the cumulative cost of all the included subsets plus the number of the uncovered items. The neighbor with the lowest cost is chosen as the starting state for the next iteration. During this local search, the subsets that are included or removed are kept track of in the tabu list for a limited number of iterations. To prevent cycles in the local search, neighbors that change the status of a subset in the tabu list are excluded from consideration.
Instance-Specific Algorithm Configuration by Yuri Malitsky