University of Cagliari

About: University of Cagliari is a education organization based out in Cagliari, Italy. It is known for research contribution in the topics: Population & Dopamine. The organization has 11029 authors who have published 29046 publications receiving 771023 citations. The organization is also known as: Università degli Studi di Cagliari & Universita degli Studi di Cagliari.

A Radio-loud Magnetar in X-ray Quiescence

Abstract: As part of a survey for radio pulsars with the Parkes 64 m telescope, we have discovered PSR J1622-4950, a pulsar with a 4.3 s rotation period. Follow-up observations show that the pulsar has the highest inferred surface magnetic field of the known radio pulsars (B {approx}3 x 10{sup 14} G), and it exhibits significant timing noise and appears to have an inverted spectrum. Unlike the vast majority of the known pulsar population, PSR J1622-4950 appears to switch off for many hundreds of days and even in its on-state exhibits extreme variability in its flux density. Furthermore, the integrated pulse profile changes shape with epoch. All of these properties are remarkably similar to the only two magnetars previously known to emit radio pulsations. The position of PSR J1622-4950 is coincident with an X-ray source that, unlike the other radio pulsating magnetars, was found to be in quiescence. We conclude that our newly discovered pulsar is a magnetar-the first to be discovered via its radio emission.

Is feature selection secure against training data poisoning

Abstract: Learning in adversarial settings is becoming an important task for application domains where attackers may inject malicious data into the training set to subvert normal operation of data-driven technologies. Feature selection has been widely used in machine learning for security applications to improve generalization and computational efficiency, although it is not clear whether its use may be beneficial or even counterproductive when training data are poisoned by intelligent attackers. In this work, we shed light on this issue by providing a framework to investigate the robustness of popular feature selection methods, including LASSO, ridge regression and the elastic net. Our results on malware detection show that feature selection methods can be significantly compromised under attack (we can reduce LASSO to almost random choices of feature sets by careful insertion of less than 5% poisoned training samples), highlighting the need for specific countermeasures.