Results 1  10
of
1,597
Comparing Predictive Accuracy
 JOURNAL OF BUSINESS AND ECONOMIC STATISTICS, 13, 253265
, 1995
"... We propose and evaluate explicit tests of the null hypothesis of no difference in the accuracy of two competing forecasts. In contrast to previously developed tests, a wide variety of accuracy measures can be used (in particular, the loss function need not be quadratic, and need not even be symmetri ..."
Abstract

Cited by 1346 (23 self)
 Add to MetaCart
We propose and evaluate explicit tests of the null hypothesis of no difference in the accuracy of two competing forecasts. In contrast to previously developed tests, a wide variety of accuracy measures can be used (in particular, the loss function need not be quadratic, and need not even
Finitetime analysis of the multiarmed bandit problem
 Machine Learning
, 2002
"... Abstract. Reinforcement learning policies face the exploration versus exploitation dilemma, i.e. the search for a balance between exploring the environment to find profitable actions while taking the empirically best action as often as possible. A popular measure of a policyâ€™s success in addressing ..."
Abstract

Cited by 817 (15 self)
 Add to MetaCart
this dilemma is the regret, that is the loss due to the fact that the globally optimal policy is not followed all the times. One of the simplest examples of the exploration/exploitation dilemma is the multiarmed bandit problem. Lai and Robbins were the first ones to show that the regret for this problem has
The Throughput of Irreducible Closed Markovian Queueing Networks: Functional Bounds, Asymptotic Loss, Efficiency, and the HarrisonWein Conjectures
, 1997
"... Let N be the population of an irreducible closed Markovian queueing network, and denote by ff u (N) the throughput of a scheduling policy u. The policy u is said to be efficient if lim N!+1 ff u (N) = ff , where ff is the capacity of a bottleneck station. The quantity J(u) := lim N!1 N(ff ..."
Abstract

Cited by 15 (6 self)
 Add to MetaCart
(ff \Gammaff u (N)) ff is called the asymptotic loss of u. The policy u is said to be asymptotically optimal if J(u) is as small as it can be. For multistation irreducible closed Markovian networks, we obtain functional bounds on the throughput which are of the form ffN N+v . The coefficients ff and v
Closed Reentrant Queueing Networks Under Affine Index Policies: Throughput Bounds, Examples and Asymptotic Loss
"... Abstract: We extend linear programming performance evaluation methods to closed reentrant queueing networks. The approach automatically generates the parameters for a surrogate of the differential cost function and enables us to obtain bounds on the system throughput at reduced computational cost th ..."
Abstract
 Add to MetaCart
the performance of the bounds and explore the asymptotic loss of the system. 1.
The asymptotic loss distribution in a fattailed factor model of portfolio credit risk. Department of Economics Working Paper 1
, 2007
"... This paper extends the standard asymptotic results concerning the percentage loss distribution in the Vasicek uniform model to a setup where the systematic risk factor is nonnormally distributed. We show that the asymptotic density in this new setup can still be obtained in closed form; in particul ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
This paper extends the standard asymptotic results concerning the percentage loss distribution in the Vasicek uniform model to a setup where the systematic risk factor is nonnormally distributed. We show that the asymptotic density in this new setup can still be obtained in closed form
Asymptotic Loss Probability in a Finite Buffer Fluid Queue with Heterogeneous HeavyTailed OnOff Processes
, 2000
"... Consider a fluid queue with a finite buffer B and capacity c fed by a superposition of N independent OnOff processes. An OnOff process consists of a sequence of alternating independent activity and silence periods. Successive activity, as well as silence, periods are identically distributed. The p ..."
Abstract

Cited by 14 (5 self)
 Add to MetaCart
the stationary loss probability and loss rate. In the case of homogeneous sources with residual activity periods equal in distribution to on r , the queue overflow probability is asymptotically, as B !1, equal to P[Q B = B] = ` N k 0 ' p k 0 on P on r ? B k 0 (r \Gamma ae) +N ae \Gamma c
Asymptotic Loss Probability in a Finite Buer Fluid Queue with Heterogeneous HeavyTailed OnO Processes
, 2000
"... Consider a
uid queue with a nite buer B and capacity c fed by a superposition of N independent OnO processes. An OnO process consists of a sequence of alternating independent activity and silence periods. Successive activity, as well as silence, periods are identically distributed. The process ..."
Abstract
 Add to MetaCart
over
ow probability and loss rate. In the case of homogeneous processes with excess activity periods equal in distribution to e, the queue loss rate is asymptotically, as B!1, equal to B
Wavelet shrinkage: asymptopia
 Journal of the Royal Statistical Society, Ser. B
, 1995
"... Considerable e ort has been directed recently to develop asymptotically minimax methods in problems of recovering in nitedimensional objects (curves, densities, spectral densities, images) from noisy data. A rich and complex body of work has evolved, with nearly or exactly minimax estimators bein ..."
Abstract

Cited by 295 (36 self)
 Add to MetaCart
Considerable e ort has been directed recently to develop asymptotically minimax methods in problems of recovering in nitedimensional objects (curves, densities, spectral densities, images) from noisy data. A rich and complex body of work has evolved, with nearly or exactly minimax estimators
REM: Active Queue Management
 IEEE NETWORK
, 2000
"... REM is an active queue management scheme that measures congestion not by a performance measure such as loss or delay, but by a quantity we call price. Price is computed by each link distributively using local information and is fed back to the sources through packet dropping or marking. This decoupl ..."
Abstract

Cited by 273 (22 self)
 Add to MetaCart
REM is an active queue management scheme that measures congestion not by a performance measure such as loss or delay, but by a quantity we call price. Price is computed by each link distributively using local information and is fed back to the sources through packet dropping or marking
reentrant lines: A computational study
, 1996
"... Asymptotic loss of priority scheduling policies in closed ..."
Results 1  10
of
1,597