Thesis (Ph. D.)--University of Texas at Austin, 2006. / Vita. Includes bibliographical references.

No description available.

We use profinite Bass-Serre theory (the theory of profinite group actions on profinite trees) to prove that the fundamental groups of finite graphs of free groups which are l-acylindrical and have finitely generated edge groups are conjugacy separable. We apply this theorem to: demonstrate that a generic positive one-relator group is conjugacy separable; produce a variant of the Rips con- struction in which the output group is conjugacy separable; apply this last to exhibit an example of a strong profinite equivalence between two finitely presented groups, one of which is conjugacy separable and the other having unsolvable conjugacy problem. We further use profinite Bass-Serre theory to demonstrate that having one end is an up-weak pro-C property for any extension- closed class C of finite groups. We show by example that it is not a down-weak pro-p property for any prime p. We consider Korenev's definition of pro-p ends for a pro-p group, and show that the number of ends of a finitely generated residually p group cannot be less than the number of pro-p ends of its pro-p completion. We explore possibilities for, but are ultimately unsuc- cessful in giving, a proper analogue of Stallings' theorem for pro-p groups. We ask which other properties might be profinite, and use another variant of the Rips construction to produce examples of patholog- ical groups such that either they are hyperbolic groups which are not residually finite, or neither property (FA) nor property (T) is an up-weak profinite property.

Let G be a semisimple linear algebraic group over an algebraically closed field of prime characteristic. In this thesis we outline the theory of such groups and their cohomology. We then concentrate on algebraic groups in rank 1 and 2, and prove some new results in their bounding cohomology.

In general, a finitely presented group can have very nasty properties, but many of these properties are avoided if the group is assumed to admit a nice action by isometries on a space with a negative curvature property, such as Gromov hyperbolicity. Such groups are surprisingly common: there is a sense in which a random group admits such an action, as do some groups of classical interest, such as fundamental groups of closed Riemannian manifolds with negative sectional curvature. If a group admits an action on a Gromov hyperbolic space then large scale properties of the space give useful invariants of the group. One particularly natural large scale property used in this way is the Gromov boundary. The Gromov boundary of a hyperbolic group is a compact metric space that is, in a sense, approximated by spheres of large radius in the Cayley graph of the group. The technical results contained in this thesis are effective versions of this statement: we see that the presence of a particular topological feature in the boundary of a hyperbolic group is determined by the geometry of balls in the Cayley graph of radius bounded above by some known upper bound, and is therefore algorithmically detectable. Using these technical results one can prove that certain properties of a group can be computed from its presentation. In particular, we show that there are algorithms that, when given a presentation for a one-ended hyperbolic group, compute Bowditch's canonical decomposition of that group and determine whether or not that group is virtually Fuchsian. The final chapter of this thesis studies the problem of detecting Cech cohomological features in boundaries of hyperbolic groups. Epstein asked whether there is an algorithm that computes the Cech cohomology of the boundary of a given hyperbolic group. We answer Epstein's question in the affirmative for a restricted class of hyperbolic groups: those that are fundamental groups of graphs of free groups with cyclic edge groups. We also prove the computability of the Cech cohomology of a space with some similar properties to the boundary of a hyperbolic group: Otal's decomposition space associated to a line pattern in a free group.

Diss--Teilw. zugl.: Univ. of Amsterdam. / Includes bibliographical references and indexes.

Thesis (M.S.)--State University of New York at Binghamton, Computer Science Department, 2006. / Includes bibliographical references.

Thesis (Ph.D.)--University of Delaware, 2008. / Principal faculty advisor: Robert Coulter, Dept. of Mathematical Sciences. Includes bibliographical references.

Project (M.S.)--Brigham Young University. Dept of Mathematics, 2006. / Includes bibliographical references (p. 19).

Let G be a simple simply connected algebraic group defined over an algebraically closed field K and V an irreducible module defined over K on which G acts. Let E denote the set of vectors in V which are eigenvectors for some non-central semisimple element of G and some eigenvalue in K*. We prove, with a short list of possible exceptions, that the dimension of Ē is strictly less than the dimension of V provided dim V > dim G + 2 and that there is equality otherwise. In particular, by considering only the eigenvalue 1, it follows that the closure of the union of fixed point spaces of non-central semisimple elements has dimension strictly less than the dimension of V provided dim V > dim G + 2, with a short list of possible exceptions. In the majority of cases we consider modules for which dim V > dim G + 2 where we perform an analysis of weights. In many of these cases we prove that, for any non-central semisimple element and any eigenvalue, the codimension of the eigenspace exceeds dim G. In more difficult cases, when dim V is only slightly larger than dim G + 2, we subdivide the analysis according to the type of the centraliser of the semisimple element. Here we prove for each type a slightly weaker inequality which still suffices to establish the main result. Finally, for the relatively few modules satisfying dim V ≤ dim G + 2, an immediate observation yields the result for dim V < dim B where B is a Borel subgroup of G, while in other cases we argue directly.