University of Portsmouth

About: University of Portsmouth is a education organization based out in Portsmouth, Portsmouth, United Kingdom. It is known for research contribution in the topics: Population & Galaxy. The organization has 5452 authors who have published 14256 publications receiving 424346 citations. The organization is also known as: Portsmouth and Gosport School of Science and Art & Portsmouth and Gosport School of Science and the Arts.

Robust ordinal regression in preference learning and ranking

Abstract: Multiple Criteria Decision Aiding (MCDA) offers a diversity of approaches designed for providing the decision maker (DM) with a recommendation concerning a set of alternatives (items, actions) evaluated from multiple points of view, called criteria. This paper aims at drawing attention of the Machine Learning (ML) community upon recent advances in a representative MCDA methodology, called Robust Ordinal Regression (ROR). ROR learns by examples in order to rank a set of alternatives, thus considering a similar problem as Preference Learning (ML-PL) does. However, ROR implements the interactive preference construction paradigm, which should be perceived as a mutual learning of the model and the DM. The paper clarifies the specific interpretation of the concept of preference learning adopted in ROR and MCDA, comparing it to the usual concept of preference learning considered within ML. This comparison concerns a structure of the considered problem, types of admitted preference information, a character of the employed preference models, ways of exploiting them, and techniques to arrive at a final ranking.

Galaxy Zoo and ALFALFA: atomic gas and the regulation of star formation in barred disc galaxies†

Abstract: We study the observed correlation between atomic gas content and the likelihood of hosting a large-scale bar in a sample of 2090 disc galaxies. Such a test has never been done before on this scale. We use data on morphologies from the Galaxy Zoo project and information on the galaxies' H I content from the Arecibo Legacy Fast Arecibo L-band Feed Array (ALFALFA) blind H I survey. Our main result is that the bar fraction is significantly lower among gas-rich disc galaxies than gas-poor ones. This is not explained by known trends for more massive (stellar) and redder disc galaxies to host more bars and have lower gas fractions: we still see at fixed stellar mass a residual correlation between gas content and bar fraction. We discuss three possible causal explanations: (1) bars in disc galaxies cause atomic gas to be used up more quickly, (2) increasing the atomic gas content in a disc galaxy inhibits bar formation and (3) bar fraction and gas content are both driven by correlation with environmental effects (e.g. tidal triggering of bars, combined with strangulation removing gas). All three explanations are consistent with the observed correlations. In addition our observations suggest bars may reduce or halt star formation in the outer parts of discs by holding back the infall of external gas beyond bar co-rotation, reddening the global colours of barred disc galaxies. This suggests that secular evolution driven by the exchange of angular momentum between stars in the bar, and gas in the disc, acts as a feedback mechanism to regulate star formation in intermediate-mass disc galaxies.

Cosmic Acceleration from Abelian Symmetry Breaking

Abstract: We discuss a consistent theory for a self-interacting vector field, breaking an Abelian symmetry in such a way to obtain an interesting behavior for its longitudinal polarization. In an appropriate decoupling limit, the dynamics of the longitudinal mode is controlled by Galileon interactions. The full theory away from the decoupling limit does not propagate ghost modes, and can be investigated in regimes where non-linearities become important. When coupled to gravity, this theory provides a candidate for dark energy, since it admits de Sitter cosmological solutions characterized by a technically natural value for the Hubble parameter. We also consider the homogeneous evolution when, besides the vector, additional matter in the form of perfect fluids is included. We find that the vector can have an important role in characterizing the universe expansion.

Probing primordial non-Gaussianity with SKA galaxy redshift surveys: a fully relativistic analysis

Abstract: The Square Kilometre Array (SKA) will produce spectroscopic surveys of tens to hundreds of millions of neutral hydrogen (H I) galaxies, eventually covering 30 000 deg2 and reaching out to redshift z≳2. The huge volumes probed by the SKA will allow for some of the best constraints on primordial non-Gaussianity, based on measurements of the large-scale power spectrum.We investigate various observational set-ups for HI galaxy redshift surveys, compatible with the SKA Phase 1 and Phase 2 (full SKA) configurations.We use the corresponding number counts and bias for each survey from realistic simulations and derive the magnification bias and the evolution of source counts directly from these. For the first time, we produce forecasts that fully include the general relativistic effects on the galaxy number counts. These corrections to the standard analysis become important on very large scales, where the signal of primordial non-Gaussianity grows strongest. Our results showthat, for the full survey, the non-Gaussianity parameter fNL can be constrained down to σ(fNL) = 1.54. This improves the current limit set by the Planck satellite by a factor of 5, using a completely different approach.

Bootstrapped Confidence Intervals as an Approach to Statistical Inference

Abstract: Confidence intervals are in many ways a more satisfactory basis for statistical inference than hypothesis tests. This article explains a simple method for using bootstrap resampling to derive confidence intervals. This method can be used for a wide variety of statistics—including the mean and median, the difference of two means or proportions, and correlation and regression coefficients. It can be implemented by an Excel spreadsheet, which is available to readers on the Web. The rationale behind the method is transparent, and it relies on almost no sophisticated statistical concepts.