Home

On the Method of Typical Bounded Differences

Abstract

Concentration inequalities are fundamental tools in probabilistic combinatorics and theoretical computer science for proving that functions of random variables are typically near their means. Of particular importance is the case where f(X) is a function of independent random variables X = (X 1, . . ., Xn ). Here the well-known bounded differences inequality (also called McDiarmid's inequality or the Hoeffding–Azuma inequality) establishes sharp concentration if the function f does not depend too much on any of the variables. One attractive feature is that it relies on a very simple Lipschitz condition (L): it suffices to show that |f(X) − f(X′)| ⩽ ck whenever X, X′ differ only in Xk . While this is easy to check, the main disadvantage is that it considers worst-case changes ck , which often makes the resulting bounds too weak to be useful.

In this paper we prove a variant of the bounded differences inequality which can be used to establish concentration of functions f(X) where (i) the typical changes are small, although (ii) the worst case changes might be very large. One key aspect of this inequality is that it relies on a simple condition that (a) is easy to check and (b) coincides with heuristic considerations as to why concentration should hold. Indeed, given an event Γ that holds with very high probability, we essentially relax the Lipschitz condition (L) to situations where Γ occurs. The point is that the resulting typical changes ck are often much smaller than the worst case ones.

To illustrate its application we consider the reverse H-free process, where H is 2-balanced. We prove that the final number of edges in this process is concentrated, and also determine its likely value up to constant factors. This answers a question of Bollobás and Erdős.

References

Hide All
[1] Achlioptas, D. (2000) Setting 2 variables at a time yields a new lower bound for random 3-SAT. In Proc. 32nd Annual ACM Symposium on Theory of Computing: STOC'00, ACM, pp. 28–37.
[2] Alon, N., Kim, J. H. and Spencer, J. (1997) Nearly perfect matchings in regular simple hypergraphs. Israel J. Math. 100 171187.
[3] Azuma, K. (1967) Weighted sums of certain dependent random variables. Tôhoku Math. J. (2) 19 357367.
[4] Bapst, V., Coja-Oghlan, A., Hetterich, S., Rassmann, F. and Vilenchik, D. The condensation phase transition in random graph coloring. Comm. Math. Phys., to appear (Proceedings version appeared in RANDOM 2014). arXiv:1404.5513.
[5] Bohman, T. (2009) The triangle-free process. Adv. Math. 221 16531677.
[6] Bohman, T., Frieze, A. and Lubetzky, E. (2015) Random triangle removal. Adv. Math. 280 379438.
[7] Bohman, T. and Keevash, P. (2010) The early evolution of the H-free process. Invent. Math. 181 291336.
[8] Bollobás, B. (1988) The chromatic number of random graphs. Combinatorica 8 4955.
[9] Bollobás, B. (2012) Personal communication.
[10] Bollobás, B. and Riordan, O. (2015) An old approach to the giant component problem. J. Combin. Theory Ser. B 113 236260.
[11] Bollobás, B. and Brightwell, G. (1992) The height of a random partial order: Concentration of measure. Ann. Appl. Probab. 2 10091018.
[12] Bollobás, B. and Riordan, O. (2009) Random graphs and branching processes. In Handbook of Large-Scale Random Networks, Vol. 18 of Bolyai Society Mathematical Studies, pp. 15–115.
[13] Borgs, C., Chayes, J. T., Mertens, S. and Pittel, B. (2004) Phase diagram for the constrained integer partitioning problem. Random Struct. Alg. 24 315380.
[14] Bushaw, N., Collares Neto, M., Morris, R. and Smith, P. (2015) The sharp threshold for maximum-size sum-free subsets in even-order abelian groups. Combin. Probab. Comput. 24 609640.
[15] Chalker, T. K., Godbole, A. P., Hitczenko, P., Radcliff, J. and Ruehr, O. G. (1999) On the size of a random sphere of influence graph. Adv. Appl. Probab. 31 596609.
[16] Chung, F. and Lu, L. (2006) Concentration inequalities and martingale inequalities: A survey. Internet Math. 3 79127.
[17] Dembo, A. and Zeitouni, O. (1998) Large Deviations Techniques and Applications, Springer.
[18] Dubhashi, D. P. and Panconesi, A. (2009) Concentration of Measure for the Analysis of Randomized Algorithms, Cambridge University Press.
[19] Erdős, P., Suen, S. and Winkler, P. (1995) On the size of a random maximal graph. Random Struct. Alg. 6 309318.
[20] Freedman, D. A. (1975) On tail probabilities for martingales. Ann. Probab. 3 100118.
[21] de Graaf, M. and Manthey, B. (2014) Probabilistic analysis of power assignments. In Mathematical Foundations of Computer Science 2014, Vol. 8635 of Lecture Notes in Computer Science, Springer, pp. 201–212.
[22] Grable, D. A. (1998) A large deviation inequality for functions of independent, multi-way choices. Combin. Probab. Comput. 7 5763.
[23] Harris, T. E. (1960) A lower bound for the critical probability in a certain percolation process. Proc. Cambridge Philos. Soc. 56 1320.
[24] Hoeffding, W. (1963) Probability inequalities for sums of bounded random variables. J. Amer. Statist. Assoc. 58 1330.
[25] Janson, S. (1990) Poisson approximation for large deviations. Random Struct. Alg. 1 221229.
[26] Janson, S., Łuczak, T. and Ruciński, A. (2000) Random Graphs, Wiley-Interscience Series in Discrete Mathematics and Optimization, Wiley-Interscience.
[27] Janson, S. and Ruciński, A. (2004) The deletion method for upper tail estimates. Combinatorica 24 615640.
[28] Janson, S. and Warnke, L. The lower tail: Poisson approximation revisited. Random Struct. Alg., to appear. arXiv:1406.1248.
[29] Kim, J. H. (1995) On Brooks' theorem for sparse graphs. Combin. Probab. Comput. 4 97132.
[30] Kim, J. H. and Vu, V. H. (2000) Concentration of multivariate polynomials and its applications. Combinatorica 20 417434.
[31] Kim, J. H. and Vu, V. H. (2004) Divide and conquer martingales and the number of triangles in a random graph. Random Struct. Alg. 24 166174.
[32] Krivelevich, M., Lubetzky, E. and Sudakov, B. (2013) Longest cycles in sparse random digraphs. Random Struct. Alg. 43 115.
[33] Makai, T. (2015) The reverse H-free process for strictly 2-balanced graphs. J. Graph Theory 79 125144.
[34] McDiarmid, C. (1989) On the method of bounded differences. In Surveys in Combinatorics, Vol. 141 of London Mathematical Society Lecture Note Series, Cambridge University Press, pp. 48188.
[35] McDiarmid, C. (1998) Concentration. In Probabilistic Methods for Algorithmic Discrete Mathematics, Vol. 16 of Algorithms and Combinatorics, Springer, pp. 195248.
[36] McKay, B. and Skerman, F. (2013) Degree sequences of random digraphs and bipartite graphs. arXiv:1302.2446.
[37] Milman, V. D. and Schechtman, G. (1986) Asymptotic Theory of Finite-Dimensional Normed Spaces, Vol. 1200 of Lecture Notes in Mathematics, Springer.
[38] Osthus, D. and Taraz, A. (2001) Random maximal H-free graphs. Random Struct. Alg. 18 6182.
[39] Riordan, O. and Warnke, L. (2015) The Janson inequalities for general up-sets. Random Struct. Alg. 46 391395.
[40] Riordan, O. and Warnke, L. (2015) The evolution of subcritical Achlioptas processes. Random Struct. Alg. 47 174203.
[41] Rödl, V. and Thoma, L. (1996) Asymptotic packing and the random greedy algorithm. Random Struct. Alg. 8 161177.
[42] Schudy, W. and Sviridenko, M. (2011) Bernstein-like concentration and moment inequalities for polynomials of independent random variables: Multilinear case. arXiv:1109.5193.
[43] Schudy, W. and Sviridenko, M. (2012) Concentration and moment inequalities for polynomials of independent random variables. In Proc. 23rd Annual ACM–SIAM Symposium on Discrete Algorithms: SODA'12, SIAM, pp. 437–446.
[44] Shamir, E. and Spencer, J. (1987) Sharp concentration of the chromatic number on random graphs Gn,p . Combinatorica 7 121129.
[45] Spencer, J. (1990) Counting extensions. J. Combin. Theory Ser. A 55 247255.
[46] Spencer, J. (1995) Asymptotic packing via a branching process. Random Struct. Alg. 7 167172.
[47] Steiger, W. L. (1969) A best possible Kolmogoroff-type inequality for martingales and a characteristic property. Ann. Math. Statist. 40 764769.
[48] Talagrand, M. (1996) A new look at independence. Ann. Probab. 24 134.
[49] Vu, V. H. (2000) On the concentration of multivariate polynomials with small expectation. Random Struct. Alg. 16 344363.
[50] Vu, V. H. (2002) Concentration of non-Lipschitz functions and applications. Random Struct. Alg. 20 262316.
[51] Warnke, L. (2014) The C -free process. Random Struct. Alg. 44 490526.
[52] Warnke, L. (2014) When does the K 4-free process stop? Random Struct. Alg. 44 355397.
[53] Wormald, N. C. (1999) Models of random regular graphs. In Surveys in Combinatorics, Vol. 267 of London Mathematical Society Lecture Note Series, Cambridge University Press, pp. 239298.

Metrics

Full text viewsFull text views reflects the number of PDF downloads, PDFs sent to Google Drive, Dropbox and Kindle and HTML full text views.

Total number of HTML views: 0
Total number of PDF views: 0 *