Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userA Multi-variate Discrimination Technique Based on Range-Searching
156
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
We present a fast and transparent multi-variate event classification technique, called PDE-RS, which is based on sampling the signal and background densities in a multi-dimensional phase space using range-searching. The employed algorithm is presented in detail and its behaviour is studied with simple toy examples representing basic patterns of problems often encountered in High Energy Physics data analyses. In addition an example relevant for the search for instanton-induced processes in deep-inelastic scattering at HERA is discussed. For all studied examples, the new presented method performs as good as artificial Neural Networks and has furthermore the advantage to need less computation time. This allows to carefully select the best combination of observables which optimally separate the signal and background and for which the simulations describe the data best. Moreover, the systematic and statistical uncertainties can be easily evaluated. The method is therefore a powerful tool to find a small number of signal events in the large data samples expected at future particle colliders.
rate research
Read More
We present a method to discriminate instanton-induced processes from standard DIS background based on Range Searching. This method offers fast and automatic scanning of a large number of variables for a combination of variables giving high signal to background ratio and the smallest theoretical and experimental uncertainties.
Between the years 2015 and 2019, members of the Horizon 2020-funded Innovative Training Network named AMVA4NewPhysics studied the customization and application of advanced multivariate analysis methods and statistical learning tools to high-energy physics problems, as well as developed entirely new ones. Many of those methods were successfully used to improve the sensitivity of data analyses performed by the ATLAS and CMS experiments at the CERN Large Hadron Collider; several others, still in the testing phase, promise to further improve the precision of measurements of fundamental physics parameters and the reach of searches for new phenomena. In this paper, the most relevant new tools, among those studied and developed, are presented along with the evaluation of their performances.
texttt{GooStats} is a software framework that provides a flexible environment and common tools to implement multi-variate statistical analysis. The framework is built upon the texttt{CERN ROOT}, texttt{MINUIT} and texttt{GooFit} packages. Running a multi-variate analysis in parallel on graphics processing units yields a huge boost in performance and opens new possibilities. The design and benchmark of texttt{GooStats} are presented in this article along with illustration of its application to statistical problems.
An observation of neutron-antineutron oscillations ($ n-bar{n}$), which violate both $B$ and $B-L$ conservation, would constitute a scientific discovery of fundamental importance to physics and cosmology. A stringent upper bound on its transition rate would make an important contribution to our understanding of the baryon asymmetry of the universe by eliminating the post-sphaleron baryogenesis scenario in the light quark sector. We show that one can design an experiment using slow neutrons that in principle can reach the required sensitivity of $tau_{n-bar{n}}sim 10^{10}s$ in the oscillation time, an improvement of $sim10^4$ in the oscillation probability relative to the existing limit for free neutrons. This can be achieved by allowing both the neutron and antineutron components of the developing superposition state to coherently reflect from mirrors. We present a quantitative analysis of this scenario and show that, for sufficiently small transverse momenta of $n/bar{n}$ and for certain choices of nuclei for the $n/bar{n}$ guide material, the relative phase shift of the $n$ and $bar{n}$ components upon reflection and the $bar{n}$ annihilation rate can be small.
Multi-variate time series (MTS) data is a ubiquitous class of data abstraction in the real world. Any instance of MTS is generated from a hybrid dynamical system and their specific dynamics are usually unknown. The hybrid nature of such a dynamical system is a result of complex external attributes, such as geographic location and time of day, each of which can be categorized into either spatial attributes or temporal attributes. Therefore, there are two fundamental views which can be used to analyze MTS data, namely the spatial view and the temporal view. Moreover, from each of these two views, we can partition the set of data samples of MTS into disjoint forecasting tasks in accordance with their associated attribute values. Then, samples of the same task will manifest similar forthcoming pattern, which is less sophisticated to be predicted in comparison with the original single-view setting. Considering this insight, we propose a novel multi-view multi-task (MVMT) learning framework for MTS forecasting. Instead of being explicitly presented in most scenarios, MVMT information is deeply concealed in the MTS data, which severely hinders the model from capturing it naturally. To this end, we develop two kinds of basic operations, namely task-wise affine transformation and task-wise normalization, respectively. Applying these two operations with prior knowledge on the spatial and temporal view allows the model to adaptively extract MVMT information while predicting. Extensive experiments on three datasets are conducted to illustrate that canonical architectures can be greatly enhanced by the MVMT learning framework in terms of both effectiveness and efficiency. In addition, we design rich case studies to reveal the properties of representations produced at different phases in the entire prediction procedure.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
