Sequential probability ratio test

About: Sequential probability ratio test is a research topic. Over the lifetime, 1248 publications have been published within this topic receiving 22355 citations.

Inference after truncated one-sided sequential test

Abstract: Assume i.i.d. random variables {X1, …, Xn, …} follow the standard exponential family {dFθ(x) = exp(θx − c(θ))dF0(x)}. For the one-sided hypothesis test H0: θ = θ0 0 where c(θ0) = c(θ1), the truncated sequential probability ratio test stops at min (τ, T) where , and H0 is rejected if τ < T. Inference problems based on asymptotic pivots are considered given τ < T by assuming T/b → finite constant larger than 1/c′(θ1).

Two-choice and three-choice probability learning in fish'

Abstract: Employing an apparatus which required a simple swimming response and a spatial discrimination, 10 African mouth-breeders were tested in two-target 80:20 and 30:70 problems, a three-target 20:60:20 problem with guidance procedure, and a 20:60:20 problem with a non-correction procedure. The sequential probability ratio test was employed in the analyses of choice data. Results showed statistically significant matching in both two-target problems, and in the three-target problems both with and without guidance procedure.

Substantiation of sequential test parameters for mass-produced electronic devices

Abstract: A planning methodology is proposed for checking the acceptability of a lot of products in terms of the percentage of defectives using the Sequential Probability Ratio Test (SPRT). For a stable, high-volume production process, in which the statistics is known (from the percentage of defectives in an accepted lot and from rejection of a whole the lot, cost per observation, the apriori probability distribution of the percentage of defectives) the sought test is to be planned, such that the expected loss due to wrong decisions and to the cost of the observations — is minimal (maximum profit). The originality of the proposed approach lies in the fact that the permissible percentage of defectives (e.g. in the form of the commonly used Acceptable Quality Level (AQL) and Limiting Quality Level (LQL)) is not specified. The only relevant information is the statistic data as mentioned above.

Using an approximate kiefer-weiss solution for testing insect population densities

Abstract: A sequential testing procedure called 2-SPRT for the meanμ of the negative binomial distribution with known exponentk is presented. For any fixedμ 0(μ 1<μ 0<μ 2), the 2-SPRT provides an asymptotic solution to the modified Kiefer-Weiss problem. Determination ofμ 0 such that the test provides an approximate solution to the Kiefer-Weiss problem is described. The behavior of the 2-SPRT and the Sequential Probability Ratio Test is investigated using Monte Carlo methods. The error probabilities and the average sample numbers are compared.

Detection method for pipeline leakage based on sequential probability ratio test

Abstract: The sequential probability ratio test (SPRT) can be applied only to the case that there is only white Guass noise in the pressure signal. Kalman filter was used to pretreatment of pressure signal to obtain the real data similar to Guass distribution. Then SPRT was used to determine leakage of pipeline. For improving the detecting accuracy of leakage positions on pipeline with SPRT, a new method based on time diversity was proposed. The deficiencies of existing pressure gradient localization method were analyzed. The experimental results proved the superiority of the proposed method.