One of the goals of machine learning is to eliminate tedious and arduous repetitive work. The manual and semi-automatic classification of millions of hours of solar wind data from multiple missions can be replaced by automatic algorithms that can discover, in mountains of multi-dimensional data, the real differences in the solar wind properties. In this paper we present how unsupervised clustering techniques can be used to segregate different types of solar wind. We propose the use of advanced data reduction methods to pre-process the data, and we introduce the use of Self-Organizing Maps to visualize and interpret 14 years of ACE data. Finally, we show how these techniques can potentially be used to uncover hidden information, and how they compare with previous manual and automatic categorizations.

We present a method based on unsupervised machine learning to identify regions of interest using particle velocity distributions as a signature pattern. An automatic density estimation technique is applied to particle distributions provided by PIC simulations to study magnetic reconnection. The key components of the method involve: i) a Gaussian mixture model determining the presence of a given number of subpopulations within an overall population, and ii) a model selection technique with Bayesian Information Criterion to estimate the appropriate number of subpopulations. Thus, this method identifies automatically the presence of complex distributions, such as beams or other non-Maxwellian features, and can be used as a detection algorithm able to identify reconnection regions. The approach is demonstrated for specific double Harris sheet simulations but it can in principle be applied to any other type of simulation and observational data on the particle distribution function.

We describe an efficient implementation of a coherent statistic for continuous gravitational wave searches from neutron stars. The algorithm works by transforming the data taken by a gravitational wave detector from a moving Earth bound frame to one that sits at the Solar System barycenter. Many practical difficulties arise in the implementation of this algorithm, some of which have not been discussed previously. These difficulties include constraints of small computer memory, discreteness of the data, losses due to interpolation and gaps in real data. This implementation is considerably more efficient than previous implementations of these kinds of searches on Laser Interferometer Gravitational Wave (LIGO) detector data.

We present direct upper limits on gravitational wave emission from the Crab pulsar using data from the first nine months of the fifth science run of the Laser Interferometer Gravitational-wave Observatory (LIGO). These limits are based on two searches. In the first we assume that the gravitational wave emission follows the observed radio timing, giving an upper limit on gravitational wave emission that beats indirect limits inferred from the spin-down and braking index of the pulsar and the energetics of the nebula. In the second we allow for a small mismatch between the gravitational and radio signal frequencies and interpret our results in the context of two possible gravitational wave emission mechanisms.

We present a method of searching for, and parameterizing, signals from known radio pulsars in data from interferometric gravitational wave detectors. This method has been applied to data from the LIGO and GEO 600 detectors to set upper limits on the gravitational wave emission from several radio pulsars. Here we discuss the nature of the signal and the performance of the technique on simulated data. We show how to perform a coherent multiple detector analysis and give some insight in the covariance between the signal parameters.

An overview of the searches for gravitational waves from radio pulsars with LIGO and GEO is given. We give a brief description of the algorithm used in these targeted searches and provide end-to-end validation of the technique through hardware injections. We report on some aspects of the recent S3/S4 LIGO and GEO search for signals from several pulsars. The gaussianity of narrow frequency bands of S3/S4 LIGO data, where pulsar signals are expected, is assessed with Kolmogorov-Smirnov tests. Preliminary results from the S3 run with a network of four detectors are given for pulsar J1939+2134.