We consider multi-label classification problems in application scenarios where classifier accu-racy is not satisfactory, but manual annotation is too costly. In single-label problems, a wellknown solution consists of using a reject option, i.e., allowing a classifier to withhold unreliabledecisions, leaving them (and only them) to human operators. We argue that this solution can be<...

General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edin...

This paper discusses structural risk minimization in the setting of classification with a reject option. Binary classification is about classifying observations that take values in an arbitrary feature space X into one of two classes, labelled −1 or +1. A discriminant function f : X → R yields a classifier sgn(f(x)) ∈ {−1,+1} that represents our guess of the label Y of a future observation X an...

This paper extends the accept reject algorithm to allow the pro posal distribution to change at each iteration We rst establish a necessary and su cient condition for this generalized accept reject al gorithm to be valid and then show how the Rao Blackwellization of Casella and Robert can be extended to this setting An impor tant application of these results is to the perfect sampling technique...

5 6 ABSTRACT. Aspiration-based evolutionary dynamics have recently been 7 used to model the evolution of fair play in the ultimatum game showing that 8 in-credible threats to reject low offers persist in equilibrium. We focus on two 9 extensions of this analysis: we experimentally test whether assumptions 10 about agent motivations (aspiration levels) and the structure of the game 11 (binary st...

Preimplantation development of the mammalian embryo consists of stages that include formation of the <span style="font-variant: normal; font-style: normal; font-family: ArialM...

The metal affinities of metal-sensing transcriptional regulators co-vary with cellular metal concentrations over more than 12 orders of magnitude. To understand the cause of this relationship, we determined the structure of the Ni(II) sensor InrS and then created cyanobacteria (Synechocystis PCC 6803) in which transcription of genes encoding a Ni(II) exporter and a Ni(II) importer were controll...

The hyperbolic frequency-modulated (HFM) waveform has an inherent Doppler-invariant property. It is more conducive than the conventional linear frequency-modulated (LFM) waveform to high speed moving target imaging. In order to apply the HFM waveform to existing inverse synthetic aperture radar (ISAR) imaging systems, a new pulse compression algorithm is proposed. First, the received HFM echoes...

Recently, Monte-Carlo Tree Search (MCTS) has advanced the field of computer Go substantially. In this article we investigate the application of MCTS for the game Lines of Action (LOA). A new MCTS variant, called MCTS-Solver, has been designed to play narrow tactical lines better in sudden-death games such as LOA. The variant differs from the traditional MCTS in respect to backpropagation and se...

Given a network together with a set of connection requests, call admission control is the problem of deciding which calls to accept and which ones to reject in order to maximize the total profit of the accepted requests. We consider call admission control problems with advance reservations in star networks. For the most general variant we present a constant-factor approximation algorithm resolv...