Apparaît dans la collection : 2016 - T1 - WS4 - Inference problems theme

Group testing is perhaps the “simplest” class of non-linear inference problems. Broadly speaking, group-testing measurements exhibit a “threshold” behaviour, with positive test outcomes if the number of items in a test are above the threshold, and negative test outcomes otherwise. In this talk we'll survey bounds, algorithms and applications for a variety of flavours of group-testing (adaptive/non-adaptive group-testing, zero-error/epsilon-error group-testing, noiseless/noisy measurements, universal group-testing, group-testing with inhibitors, constrained group-testing). The talk is intended as a survey of classical and recent work, and will also present some open questions.