How close can we come to a parity function when there isn't one?
نویسندگان
چکیده
Consider a group G such that there is no homomorphism f : G → {±1}. In that case, how close can we come to such a homomorphism? We show that if f has zero expectation, then the probability that f(xy) = f(x)f(y), where x, y are chosen uniformly and independently from G, is at most 1/2(1 + 1/ √ d), where d is the dimension of G’s smallest nontrivial irreducible representation. For the alternating group An, for instance, d = n − 1. On the other hand, An contains a subgroup isomorphic to Sn−2, whose parity function we can extend to obtain an f for which this probability is 1/2(1 + 1/ ( n 2 ) ). Thus the extent to which f can be “more homomorphic” than a random function from An to {±1} lies between O(n) and Ω(n). The symmetric group Sn has a parity function, i.e., a homomorphism f : Sn → {±1}, sending even and odd permutations to +1 and −1 respectively. The alternating group An, which consists of the even permutations, has no such homomorphism. How close can we come to one? What is the maximum, over all functions f : An → {±1} with zero expectation, of the probability Pr x,y [f(x)f(y) = f(xy)] , where x and y are chosen independently and uniformly from An? We give simple upper and lower bounds on this quantity, for groups in general and for An in particular. Our results are easily extended to functions f : G → C, but we do not do this here. Our main result is the following: Theorem 1. Let G be a group, and let f : G → {±1} such that Ef = 0. Then Pr x,y [f(x)f(y) = f(xy)] ≤ 1 2 ( 1 + 1 √ d ) , where d = minρ6=1 dρ is the dimension of the smallest nontrivial irreducible representation of G. Thus if G is quasirandom in Gowers’ sense [1]—that is, if minρ6=1 dρ is large—it is impossible for f to be much more homomorphic than a uniformly random function. For An in particular, the dimension of the smallest nontrivial representation is d = n−1, so Prx,y[f(x)f(y) = f(xy)]−1/2 = O(1/ √ n). If f is a class function, i.e., if f is invariant under conjugation so that f(xyx) = f(y) for all x, y ∈ G, then we can tighten this bound from 1/ √ d to 1/d:
منابع مشابه
Climate Virtues Ethics: A Proposal for Future Research
Climate virtue ethics points to the subjective/personal dimensions of climate ethics, which have been largely neglected by previous research. There is a lot of research from diverse fields that pertains to the cultural and the individual dimensions that come along with climate virtue ethics, but, as of yet, these dimensions have hardly been examined together. Future research on climate virtue e...
متن کاملمدیر موفق کیست؟
Who is a really successful manager? A manager who spends less money, or the one who earns more? A manager who can survive for a longer period of time, or an administrator who expands his organization, and opens up new branches? Which one is the most successful? The article tries to answer these questions and provides, some simple guidlines for the managers in every domain of management who wan...
متن کاملThe Structure of Bhattacharyya Matrix in Natural Exponential Family and Its Role in Approximating the Variance of a Statistics
In most situations the best estimator of a function of the parameter exists, but sometimes it has a complex form and we cannot compute its variance explicitly. Therefore, a lower bound for the variance of an estimator is one of the fundamentals in the estimation theory, because it gives us an idea about the accuracy of an estimator. It is well-known in statistical inference that the Cram&eac...
متن کاملThree questions you need to ask about your brand.
Traditionally, the people responsible for positioning brands have concentrated on the differences that set each brand apart from the competition. But emphasizing differences isn't enough to sustain a brand against competitors. Managers should also consider the frame of reference within which the brand works and the features the brand shares with other products. Asking three questions about your...
متن کاملروش انتگرال مسیر برای مدل هابارد تک نواره
We review various ways to express the partition function of the single-band Hubard model as a path integral. The emphasis is made on the derivation of the action in the integrand of the path integral and the results obtained from this approach are discussed only briefly. Since the single-band Hubbard model is a pure fermionic model on the lattice and its Hamiltonian is a polynomial in creat...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- CoRR
دوره abs/1005.4897 شماره
صفحات -
تاریخ انتشار 2010