Inferences for case-control and semiparametric two-sample density ratio models

By Y. Qin. 1998

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections (2019, methodology): “…various direct ratio estimators have been proposed [44], including the moment matching approach [17], probabilistic classiﬁcation approach [4, 7, 40], and ratio matching approach [22, 33, 45] The proposed DualDICE algorithm, as a direct approach for density ratio estimation, bears some…”
Learning in Implicit Generative Models (2016, methodology): “[19] ca ll this approach unsupervised-as-supervised learning, Qin [31] explore this for analysis of case-contro l in observational studies, both Neal [28] and Cranmer et al.”
Likelihood-Free Inference by Ratio Estimation (2016, methodology): “, 2014), of which estimation through logistic regression is widely used and has some favourable asymptotic properties (Geyer, 1994; Qin, 1998; Cheng and Chu, 2004; Bickel et al., 2007).”
Least-squares two-sample test (2011, methodology): “Thanks to the analytic-form expression, uLSIF is computationally more efficient than alternative density ratio estimators which involve non-linear optimization (Qin, 1998; Cheng & Chu, 2004; Huang et al., 2007; Sugiyama et al., 2008; Nguyen et al., 2010).”
On empirical likelihood for semiparametric two-sample density ratio models (2008, methodology): “(a) We prove this part using some similar arguments to those in Qin and Lawless (1994) and Zou et al.”
Merging information for semiparametric density estimation (2004, methodology): “The asymptotic theory generalizes the theory of both Qin (1998) and Fokianos et al. (2001) in the sense that it covers the more general case of m samples with weights not of the exponential form necessarily.”
Generative Adversarial Nets from a Density Ratio Estimation Perspective (2016, background): “17 derived from the logistic model is known to be optimal [33].”
Under Review as a Conference Paper at Iclr 2017 Learning in Implicit Generative Models (2017, background): “Hastie et al. (2013) call this approach unsupervised-assupervised learning, Qin (1998) explore this for analysis of case-control in observational studies, both Neal (2008) and Cranmer et al. (2015) explore this approach for high-energy physics applications, Gutmann and Hyvärinen (2012) exploit it…”
A Least-squares Approach to Direct Importance Estimation (2009, methodology): “…we can formulate our optimization problem as follows. min α∈Rb [ 1 2 α⊤Ĥα− ĥ⊤α+λ1⊤b α ] subject toα≥ 0b, (6) where 0b and 1b are theb-dimensional vectors with all zeros and ones, respectively; the vector inequalityα≥ 0b is applied in the element-wise manner, that is,αℓ≥ 0 for ℓ = 1,2, . . . ,b.”
Conditional Density Estimation via Least-Squares Density Ratio Estimation (2010, methodology): “The kernel mean matching method (Huang et al., 2007) and the logistic regression based method (Qin, 1998; Cheng & Chu, 2004; Bickel et al., 2007) also allow one to directly estimate a density ratio q(x)/q′(x).”
Fair Transfer Learning with Missing Protected Attributes (2019, methodology): “Logistic regression does this naturally and has been shown to be optimal (minimum asymptotic variance) for correctly specified models [19] but performs poorly for mis-specified models [25].”
Density-ratio matching under the Bregman divergence: a unified framework of density-ratio estimation (2012, methodology): “Below, we consider the induction setup, where the entire density-ratio function r∗(x) is estimated (Qin 1998; Kanamori et al. 2012).”
Wild Variational Approximations (2016, methodology): “A new direction for energy approximation applies density ratio estimation methods [33, 34, 35].”
Testing for a finite mixture model with two components (2004, background): “ture models. Some important more recent developments can be found in Cheng and Traylor (1995), Bickel and Cherno (1993), Cherno and Lander (1995), Lemdani and Pons (1999), Dacunha-Castelle and Gassiat (1999) and Chen and Chen (2001), Qin (1998) , Liang and Rathouz (1999).”
Fairness-Aware Learning with Restriction of Universal Dependency using f-Divergences (2015, methodology): “[13] presented a divergence estimator of the f -divergences based on using the moment matching estimator [17] to estimate the probability ratio.”
A theory of statistical models for Monte Carlo integration (2003, background): “Let n = n1 + n2 and denote the pooled sample .T1, ::: , Tn/ = .X1, ::: , Xn1 ; Z1, ::: , Zn2 /. Following the results of Anderson (1979) and Qin (1998) , inferences for the vector .α, θ/ can be based on the profile log-likelihood”