Abstract
We consider a simple model of imprecise comparisons: there exists some δ > 0 such that when a subject is given two elements to compare, if the values of those elements (as perceived by the subject) differ by at least δ, then the comparison will be made correctly; when the two elements have values that are within δ, the outcome of the comparison is unpredictable. This model is inspired by both imprecision in human judgment of values and also by bounded but potentially adversarial errors in the outcomes of sporting tournaments. Our model is closely related to a number of models commonly considered in the psychophysics literature where δ corresponds to the Just Noticeable Difference (JND) unit or difference threshold. In experimental psychology, the method of paired comparisons was proposed as a means for ranking preferences among n elements of a human subject. The method requires performing all (n2) comparisons, then sorting elements according to the number of wins. The large number of comparisons is performed to counter the potentially faulty decision-making of the human subject, who acts as an imprecise comparator. We show that in our model the method of paired comparisons has optimal accuracy, minimizing the errors introduced by the imprecise comparisons. However, it is also wasteful because it requires all (n2). We show that the same optimal guarantees can be achieved using 4n3/2 comparisons, and we prove the optimality of our method. We then explore the general tradeoff between the guarantees on the error that can be made and number of comparisons for the problems of sorting, max-finding, and selection. Our results provide strong lower bounds and close-to-optimal solutions for each of these problems.
- Gagan Aggarwal, Nir Ailon, Florin Constantin, Eyal Even-Dar, Jon Feldman, Gereon Frahling, Monika R. Henzinger, S. Muthukrishnan, Noam Nisan, Martin Pál, Mark Sandler, and Anastasios Sidiropoulos. 2008. Theory research at Google. SIGACT News 39, 2 (2008), 10--28. Google ScholarDigital Library
- Noga Alon and Yossi Azar. 1988. Sorting, approximate sorting, and searching in rounds. SIAM Journal on Discrete Math 1, 3 (1988), 269--280. Google ScholarDigital Library
- Shay Assaf and Eli Upfal. 1991. Fault tolerant sorting networks. SIAM Journal on Discrete Math 4, 4 (1991), 472--480. Google ScholarDigital Library
- Michael Ben-Or and Avinatan Hassidim. 2008. The bayesian learner is optimal for noisy binary search (and pretty good for quantum as well). In Proceedings of the 49th Annual IEEE Symposium on Foundations of Computer Science (FOCS). 221--230. Google ScholarDigital Library
- Manuel Blum, Robert W. Floyd, Vaughan R. Pratt, Ronald L. Rivest, and Robert E. Tarjan. 1973. Time bounds for selection. Journal of Computer and Systems Science 7, 4 (1973), 448--461. Google ScholarDigital Library
- Béla Bollobás and Andrew Thomason. 1983. Parallel sorting. Discrete Applied Mathematics 6 (1983), 1--11.Google ScholarCross Ref
- Ryan S. Borgstrom and S. Rao Kosaraju. 1993. Comparison-based search in the presence of errors. In Proceedings of the 25th Annual ACM Symposium on Theory of Computing (STOC). 130--136. Google ScholarDigital Library
- Ferdinando Cicalese. 2013. Fault-Tolerant Search Algorithms - Reliable Computation with Unreliable Information. Springer. 1--198 pages. Google ScholarDigital Library
- Herbert Aron David. 1988. The Method of Paired Comparisons (2nd ed.). Charles Griffin & Company Limited.Google Scholar
- Uriel Feige, Prabhakar Raghavan, David Peleg, and Eli Upfal. 1994. Computing with noisy information. SIAM Journal on Computing 23, 5 (1994). Google ScholarDigital Library
- Vitaly Feldman. 2009. Robustness of evolvability. In Proceedings of COLT. 277--292.Google Scholar
- Irene Finocchi, Fabrizio Grandoni, and Giuseppe F. Italiano. 2009. Optimal resilient sorting and searching in the presence of memory faults. Theoretical Computer Science 410 (2009), 4457--4470. Issue 44. Google ScholarDigital Library
- Irene Finocchi and Giuseppe F. Italiano. 2004. Sorting and searching in the presence of memory faults (without redundancy). In Proceedings of the 36th Annual ACM Symposium on Theory of Computing (STOC). 101--110. Google ScholarDigital Library
- William I. Gasarch, Evan Golub, and Clyde P. Kruskal. 2003. Constant time parallel sorting: An empirical view. Journal of Computer and System Science 67, 1 (2003), 63--91. Google ScholarDigital Library
- Roland Häggkvist and Pavol Hell. 1981. Parallel sorting with constant time for comparisons. SIAM Journal on Computing 10, 3 (1981), 465--472.Google ScholarCross Ref
- Roland Häggkvist and Pavol Hell. 1982. Sorting and merging in rounds. SIAM Journal on Algebraic Discrete Methods 3, 4 (1982), 465--473.Google ScholarCross Ref
- Richard M. Karp and Robert Kleinberg. 2007. Noisy binary search and its applications. In SODA. 881--890. Google ScholarDigital Library
- Michael Kearns. 1998. Efficient noise-tolerant Learning from statistical queries. Journal of the ACM 45, 6 (1998), 983--1006. Google ScholarDigital Library
- H. G. Landau. 1953. On dominance relations and the structure of animal societies: III. The condition for a score structure. The Bulletin of Mathematical Biophysics 15, 2 (1953), 143--148. DOI:http://dx.doi.org/10.1007/BF02476378Google ScholarCross Ref
- László Lovász. 1986. An Algorithmic Theory of Numbers, Graphs, and Convexity. CBMS-NSF Regional Conference Series in Applied Mathematics 50, SIAM.Google ScholarCross Ref
- Stephen B. Maurer. 1980. The king chicken theorems. Mathematics Magazine 53, 2 (March 1980), 67--80.Google ScholarCross Ref
- Andrzej Pelc. 2002. Searching games with errors—fifty years of coping with liars. Theoretical Computer Science 270, 1--2 (2002), 71--109. Google ScholarDigital Library
- Bala Ravikumar, K. Ganesan, and K. B. Lakshmanan. 1987. On selecting the largest element in spite of erroneous information. In Proceedings of the 4th Annual Symposium on Theoretical Aspects of Computer Science (STACS). 88--99. Google ScholarDigital Library
- Alfréd Rényi. 1962. On a problem in information theory. Magyar Tud. Akad. Mat. Kutató Int. Közl 6 (1962), 505--516.Google Scholar
- Ronald L. Rivest, Albert R. Meyer, Daniel J. Kleitman, Karl Winklmann, and Joel Spencer. 1980. Coping with errors in binary search procedures. Journal of Computer and System Science 20, 3 (1980), 396--405.Google ScholarCross Ref
- Jian Shen, Li Sheng, and Jie Wu. 2003. Searching for sorted sequences of kings in tournaments. SIAM Journal on Computing 32, 5 (2003), 1201--1209. Google ScholarDigital Library
- Scott M. Smith and Gerald S. Albaum. 2005. Fundamentals of Marketing Research (1st ed.). Sage Publications, Inc.Google Scholar
- Louis Leon Thurstone. 1927. A law of comparative judgment. Psychological Review 34 (1927), 273--286.Google ScholarCross Ref
- Stanisław Marcin Ulam. 1976. Adventures of a Mathematician. Scribner’s, New York.Google Scholar
- Leslie G. Valiant. 1975. Parallelism in comparison problems. SIAM Journal on Computing 4, 3 (1975), 348--355.Google ScholarDigital Library
- Andrew Chi-Chih Yao and Frances Foong Yao. 1985. On fault-tolerant networks for sorting. SIAM Journal on Computing 14, 1 (1985), 120--128.Google ScholarCross Ref
Index Terms
- Sorting and Selection with Imprecise Comparisons
Recommendations
Sorting and Selection with Imprecise Comparisons
ICALP '09: Proceedings of the 36th International Colloquium on Automata, Languages and Programming: Part IIn experimental psychology, the method of paired comparisons was proposed as a means for ranking preferences amongst n elements of a human subject. The method requires performing all $\binom{n}{2}$ comparisons then sorting elements according to the ...
Sizing user stories using paired comparisons
Agile estimation approaches usually start by sizing the user stories to be developed by comparing them to one another. Various techniques, with varying degrees of formality, are used to perform the comparisons - plain contrasts, triangulation, planning ...
Energy Efficient Sorting, Selection and Searching
WALCOM: Algorithms and ComputationAbstractIn this paper, we introduce a model for studying energy efficient algorithms by extending the well-studied comparison model. In our model, the result of a comparison is determined based on two parameters: (i) the energy used to perform a ...
Comments