Reasoning About Approximate Match Query Results

Author(s):  
S. Guha ◽  
N. Koudas ◽  
D. Srivastava ◽  
Xiaohui Yu
Keyword(s):  
Author(s):  
Tetsushi Matsui ◽  
Takeaki Uno ◽  
Juzoh Umemori ◽  
Tsuyoshi Koide

2020 ◽  
Vol 7 ◽  
Author(s):  
Wei Quan ◽  
Dengfeng Guan ◽  
Guangri Quan ◽  
Bo Liu ◽  
Yadong Wang

Author(s):  
Sudipto Guha ◽  
Nick Koudas ◽  
Amit Marathe ◽  
Divesh Srivastava
Keyword(s):  

2020 ◽  
pp. 75-107
Author(s):  
Paul Noordhof

The semantics for counterfactuals builds on Lewis’ theory and similarity weighting, with two principal aims. The first of these is to resist those who argue that appeal to counterfactuals in the analysis of causation is circular because their semantics must appeal to causal facts (specifically, in the similarity weighting for worlds). The second of these is to secure the semantics for counterfactuals for a possible account of causal non-symmetry in terms of counterfactual non-symmetry. As a result, weight is rightly placed upon perfect match as an important aspect of similarity although this is moderated to deal an indeterministic version of Fine’s Future Similarity Objection. A more precise understanding of the role of approximate match is developed appealing to a type of probabilistic independence.


1995 ◽  
Vol 15 (5) ◽  
pp. 951-960 ◽  
Author(s):  
Katalin Marton ◽  
Paul C. Shields

AbstractAlmost-sure convergence of (l/k) log Wk(x, y) to entropy for weak Bernoulli processes is proved, where Wk (x, y) is the waiting time until an initial segment of length k of a sample path x is seen in an independently chosen sample path y. Analogous almost-sure results are obtained in the approximate match case for very weak Bernoulli processes. The weak Bernoulli proof uses recent results obtained by the authors about the estimation of joint distributions, while the very weak Bernoulli result utilizes a new characterization of such processes in terms of a blowing-up property.


Sign in / Sign up

Export Citation Format

Share Document