Improving survey specifications are causing an exponential rise in pulsar candidate numbers and data volumes. We study the candidate filters used to mitigate these problems during the past fifty years. We find that some existing methods such as applying constraints on the total number of candidates collected per observation, may have detrimental effects on the success of pulsar searches. Those methods immune to such effects are found to be ill-equipped to deal with the problems associated with increasing data volumes and candidate numbers, motivating the development of new approaches. We therefore present a new method designed for on-line operation. It selects promising candidates using a purpose-built tree-based machine learning classifier, the Gaussian Hellinger Very Fast Decision Tree (GH-VFDT), and a new set of features for describing candidates. The features have been chosen so as to i) maximise the separation between candidates arising from noise and those of probable astrophysical origin, and ii) be as survey-independent as possible. Using these features our new approach can process millions of candidates in seconds (∼1 million every 15 seconds), with high levels of pulsar recall (90%+). This technique is therefore applicable to the large volumes of data expected to be produced by the Square Kilometre Array (SKA). Use of this approach has assisted in the discovery of 20 new pulsars in data obtained during the LOFAR Tied-Array All-Sky Survey (LOTAAS).

A B S T R A C TWe study a mean field model of the solar dynamo, in which the non-linearity is the action of the azimuthal component of the Lorentz force of the dynamo-generated magnetic field on the angular velocity. The underlying zero-order angular velocity is consistent with recent determinations of the solar rotation law, and the form of the alpha effect is chosen so as to give a plausible butterfly diagram. For small Prandtl numbers we find regular, intermittent and apparently chaotic behaviour, depending on the size of the alpha coefficient. For certain parameters, the intermittency displays some of the characteristics believed to be associated with the Maunder minimum. We thus believe that we are capturing some features of the solar dynamo.

Abstract. We study a Cartesian analogy for a solar dynamo model to investigate systematically a dynamo model limited by the back reaction of the Lorentz force on the differential rotation. In particular, we investigate intermittent behaviour found at low turbulent magnetic Prandtl numbers τ, and determine empirical scaling laws with τ. We find this class of models to be incapable of producing extended periods of "normal" behaviour separated by occasional "grand minima" -rather the behaviour is a mirror image, with occasional "grand maxima". Further we find the existence of solar-like torsional oscillations to be incompatible with low magnetic Prandtl number intermittent regimes.

Abstract-Classifiers trained on data sets possessing an imbalanced class distribution are known to exhibit poor generalisation performance. This is known as the imbalanced learning problem. The problem becomes particularly acute when we consider incremental classifiers operating on imbalanced data streams, especially when the learning objective is rare class identification. As accuracy may provide a misleading impression of performance on imbalanced data, existing stream classifiers based on accuracy can suffer poor minority class performance on imbalanced streams, with the result being low minority class recall rates. In this paper we address this deficiency by proposing the use of the Hellinger distance measure, as a very fast decision tree split criterion. We demonstrate that by using Hellinger a statistically significant improvement in recall rates on imbalanced data streams can be achieved, with an acceptable increase in the false positive rate.

Context. It has recently been claimed that analysis of Greenwich sunspot data over 120 years reveals that sunspot activity clusters around two longitudes separated by 180• ("active longitudes") with clearly defined differential rotation during activity cycles. In previous work we demonstrated that such effects can be observed in synthetic data without such features, as an artefact of the method of analysis. Aims. In the present work we extend this critical examination of methodology to the actual Greenwich sunspot data and also consider newly proposed methods of analysis claiming to confirm the original identification of active longitudes. Methods. We performed fits of different kinematic frames onto the actual sunspot data. Firstly, a cell-counting statistic was used to analyse a comoving system of frames and show that such frames extract useful information from the data. Secondly, to check the claim of century-scale persistent active longitudes in a contramoving frame system, we made a comprehensive exploration of parameter space following the original methodology as closely as possible. Results. Our analysis revealed that values obtained for the parameters of differential rotation are not stable across different methods of analysis proposed to track persistent active longitudes. Also, despite a very thorough search in parameter space, we were unable to reproduce results claiming to reveal the century-persistent active longitudes. Previous parameter space exploration has been restricted to frames whose latitudinal profile is opposite to solar surface differential rotation. Relaxing this restriction we found that the highest values of nonaxisymmetry occur for frames comoving with the solar surface flow. Further analysis indicates that even these solutions are the result of purely statistical fluctuations. Conclusions. We can therefore say that strong and well substantiated evidence for an essential and century-scale persistent nonaxisymmetry in the sunspot distribution does not exist.