Results 41 to 50 of about 26,576 (302)
The Heavy-Tail Phenomenon in SGD
CoRR, 2020 In recent years, various notions of capacity and complexity have been proposed for characterizing the generalization properties of stochastic gradient descent (SGD) in deep learning. Some of the popular notions that correlate well with the performance on unseen data are (i) the `flatness' of the local minimum found by SGD, which is related to the ...
Gurbuzbalaban, Mert +2 more
openaire +4 more sources
Bayesian Robustness Modelling Using the Floor Distribution
Revstat Statistical Journal, 2018 In Bayesian analysis, the prior distribution and the likelihood can conflict, in the sense that they can carry diverse information about the parameter of interest. The most common form of conflict is the presence of outliers in the data.
J. A. A. Andrade +2 more
doaj +1 more source
A Review of More than One Hundred Pareto-Tail Index Estimators
Statistica, 2021 Heavy-tailed distributions are often encountered in economics, finance, biology, telecommunications, geology, etc. The heaviness of a tail is measured by a tail index. Numerous methods for tail index estimation have been proposed. This paper reviews more
Igor Fedotenkov
doaj +1 more source
Heavy Tails in Program Structure
IEEE Computer Architecture Letters, 2017 Designing and optimizing computer systems require deep understanding of the underlying system behavior. Historically many important observations that led to the development of essential hardware and software optimizations were driven by empirical observations about program behavior. In this paper, we report an interesting property of program structures
Hiroshi Sasaki 0001 +3 more
openaire +1 more source
Heavy Metals: Might as Well Jump
International Journal of Financial Studies, 2019 Financial times series, and commodity prices in particular, are known to exhibit fat tails in the distribution of prices. As with many natural resources price series, the arrival of new information can lead to unexpectedly rapid changes—or jump ...
Neil A. Wilmot
doaj +1 more source
The Heavy Tails of Vulnerability Exploitation [PDF]
, 2015 In this paper we analyse the frequency at which vulnerabilities are exploited in the wild by relying on data collected worldwide by Symantec’s sensors. Our analysis comprises 374 exploited vulnerabilities for a total of 75.7 Million recorded attacks spanning three years (2009-2012).
openaire +1 more source
The heavy tail of the human brain
Current Opinion in Neurobiology, 2015 Fluctuating oscillations are a ubiquitous feature of neurophysiology. Are the amplitude fluctuations of neural oscillations chance excursions drawn randomly from a normal distribution, or do they tell us more? Recent empirical research suggests that the occurrence of 'anomalous' (high amplitude) oscillations imbues their probability distributions with ...
Roberts, James A +2 more
openaire +4 more sources
The Scientific World Journal, 2015 I present a parametric, bijective transformation to generate heavy tail versions of arbitrary random variables. The tail behavior of this heavy tail Lambert W × FX random variable depends on a tail parameter δ≥0: for δ=0, Y≡X, for δ>0 Y has heavier ...
Georg M. Goerg
doaj +1 more source
Molecular Oncology, EarlyView.We have established a humanized orthotopic patient‐derived xenograft (Hu‐oPDX) mouse model of high‐grade serous ovarian cancer (HGSOC) that recapitulates human tumor–immune interactions. Using combined anti‐PD‐L1/anti‐CD73 immunotherapy, we demonstrate the model's improved biological relevance and enhanced translational value for preclinical ...
Luka Tandaric +10 more
wiley +1 more source
A test procedure for detecting super-heavy tails [PDF]
, 2009 The aim of this work is to develop a test to distinguish between heavy and super-heavy tailed probability distributions. These classes of distributions are relevant in areas such as telecommunications and insurance risk, among others.
de Haan, L.F.M. +5 more
core +1 more source