Search
Search Funnelback University
- Refined by:
- Date: 2013
1 -
20 of
25
search results for news |u:www.statslab.cam.ac.uk
Fully-matching results
-
Statistics
www.statslab.cam.ac.uk/~rrw1/stats/index.html31 Jul 2013: What is here is still relevant, but the new schedule covers the general linear model, and Gauss-Markov theorem, which is not covered in these notes. ... Chance News reviews current issues in the news that use probability or statistical concepts. -
Fixed-point models
www.statslab.cam.ac.uk/~frank/PAPERS/fpmee.html10 May 2013: Abstract. This paper presents a new approach to modeling end-to-end performance for IP networks. ... Results are presented for an IP backbone network, which highlight how this new model finds the natural operating point for TCP, which depends on route -
Example sheet 1 1 Conditional expectation Exercise 1.1. Let ...
www.statslab.cam.ac.uk/~ps422/examples1.pdf16 Oct 2013: Ateach time 1, 2, 3,. , a ball is chosen at random from the urn and is replaced together witha new ball of the same colour. ... Just before each time n = 1, 2,. , a new gambler arrives on the scene. -
WIAS2013.dvi
www.statslab.cam.ac.uk/~rjs57/WIAS2013.pdf3 Feb 2013: LOG-CONCAVITY: NEW THEORY AND METHODOLOGY. Co-authors: Y. Chen, M. Cule, L. ... Sinica, to appear. • Comon, P. (1994) Independent component analysis, A new conc ept? -
Self-avoiding walksand the Fisher transformation Geoffrey Grimmett…
www.statslab.cam.ac.uk/~grg/papers/fisher5.pdf11 Sep 2013: The transformation maps G to a new graph F(G). We have two sets of results.First, the connective constants of G and F(G) satisfy a simple functional relation, and -
Richard Weber's Publications
www.statslab.cam.ac.uk/~rrw1/publications/papers.html8 Nov 2013: In Proc. 23 Annual ACM Symposium on Theory of Computing, New Orleans, May 6-8, pages 230-240, 1991. ... on Theory of Computing, pages 208213. ACM, New York, 2000. -
Discussion of ‘Correlated variables in regression: clustering and…
www.statslab.cam.ac.uk/~rds37/papers/Shah%20Samworth%202013%20Discussion%20of%20correlated%20variables%20in%20regression%20-%20clustering%20and%20sparse%20estimation.pdf18 Jan 2013: new direction suggested by the work. -
QLE Jason Miller and Scott Sheffield MIT August 1, ...
www.statslab.cam.ac.uk/~jpm205/slides/qleslides.pdf1 Aug 2013: Can we generalize DLA and FPP?I When FPP weights are exponential, growth process selects new edges from. -
Nonparametric Bernstein-von Mises theorems in Gaussian white noise
www.statslab.cam.ac.uk/~nickl/Site/__files/AOS1133.pdf28 Oct 2013: The Annals of Statistics2013, Vol. 41, No. 4, 1999–2028DOI: 10.1214/13-AOS1133 Institute of Mathematical Statistics, 2013. NONPARAMETRIC BERNSTEIN–VON MISES THEOREMS INGAUSSIAN WHITE NOISE. BY ISMAËL CASTILLO1 AND RICHARD NICKL. CNRS and -
A permuted random walk exits faster Richard Pymar∗ Perla ...
www.statslab.cam.ac.uk/~ps422/walkfinal.pdf25 Apr 2013: In view of (2.1) in the new graph indegG(i) = indegG′(i) = outdegG′(i). ... Springer-Verlag, New York, 1998. [3] A. Burchard and M. Schmuckenschläger. Comparison theorems for exit times. -
EXTENDABLE SELF-AVOIDING WALKS GEOFFREY R. GRIMMETT, ALEXANDER E.…
www.statslab.cam.ac.uk/~grg/papers/ext22.pdf23 Nov 2013: In both cases, each thickened portion of w isbackward extendable. so that the old wa becomes the new w0.) Since loop-erasure does notincrease distances along the walk among the vertices -
seminar.dvi
www.statslab.cam.ac.uk/~rrw1/talks/mechdesignseminar.pdf28 Feb 2013: Let new fees be as follows. p1(θ) = p1(θ1) x(θ) E[. ... Then adopt new payments:. p1(θ1, θ2, θ3,. , θn) = p1(θ1, θ2, θ3,. , -
Mixing times are hitting times of large sets Yuval ...
www.statslab.cam.ac.uk/~ps422/mix-hit.pdf25 Apr 2013: In this paper we establish a new connection between mixing times and hittingtimes for reversible Markov chains (Theorem 1.1). ... 8 A new proof of tprod tL for reversible chains. Recall the definition tprod = maxx,A. -
Universality for bond percolation in two dimensions
www.statslab.cam.ac.uk/~grg/papers/AOP740.pdf12 Sep 2013: 1987). Tilings and Patterns. Freeman, New York.MR0857454. [8] KENYON, R. (2004). ... Springer, New York.MR0894549. [12] KESTEN, H. (1987). Scaling relations for 2D-percolation. -
The Multi-Armed Bandit Problem: Index Theory Since Gittins Richard ...
www.statslab.cam.ac.uk/~rrw1/talks/gocps.pdf7 Feb 2013: is about to quit —, we increase the prize tog(y), which becomes the new ‘prevailing prize’. ... is about to quit —, we increase the prize tog(y), which becomes the new ‘prevailing prize’. • -
Dimension of Fractional Brownian motion with variable drift Yuval ...
www.statslab.cam.ac.uk/~ps422/dim-graph-image.pdf30 Oct 2013: It remains to show that I2 <. By defininga new equivalence relation on rectangles in Dk, i.e. ... Springer-Verlag, New York, 1995. [10] D. Khoshnevisan and Y. Xiao. Brownian motion and thermal capacity. -
PERCOLATION OF FINITE CLUSTERSAND INFINITE SURFACES GEOFFREY R.…
www.statslab.cam.ac.uk/~grg/papers/plaq-final.pdf7 Mar 2013: By Lemma 4.4again, X(ω′) X(ω) is finite, so no new infinite components have beencreated, and thus X(ω′) has strictly fewer that n infinite components.Since the modification -
Tutorial Bandit Processes and Index Policies Richard Weber,…
www.statslab.cam.ac.uk/~rrw1/talks/YETQweber2013.pdf14 Nov 2013: K. D. Glazebrook, D. J. Hodge, C. Kirkbride, R. J. Minty, Stochastic scheduling: Ashort history of index policies and new approaches to index generation for dynamicresource allocation, J Scheduling., 2013. ... Dating strategy: should I contact a new -
Inhomogeneous bond percolation on square, triangular and hexagonal…
www.statslab.cam.ac.uk/~grg/papers/AOP729.pdf3 Jul 2013: It implies criticality of such values, therebyproviding a new proof of the critical point of inhomogeneous systems. ... A great deal of rigorous mathematics exists for critical site percolation on thetriangular lattice, but surprisingly little for other -
Advanced Probability Perla Sousi∗ December 17, 2011 Contents 1 ...
www.statslab.cam.ac.uk/~ps422/notes-2011.pdf10 Oct 2013: It is easy to check thatG = {iJBi : J N}.The natural thing to do is to define a new random variable X′ = E[X|G] as follows. ... So for allA Fn we have. E[Xn1(A)] = E[X1(A)]. Hence if we now define a new probability measure Q̃(A) = E[X1(A)], then Q(A) =
Search history
Recently clicked results
Recently clicked results
Your click history is empty.
Recent searches
Recent searches
Your search history is empty.