Marco Frailis

Self-organizing networks for classification: developing applications to science analysis for astroparticle physics

Physics analysis in astroparticle experiments requires the capability of recognizing new phenomena; in order to establish what is new, it is important to develop tools for automatic classification, able to compare the final result with data from different detectors. A typical example is the problem of gamma ray burst detection, classification, and possible association to known sources: for this task physicists will need in the next years tools to associate data from optical databases, from satellite experiments (EGRET, GLAST), and from Cherenkov telescopes (MAGIC, HESS, CANGAROO, VERITAS).

The Use of HepRep in GLAST

HepRep is a generic, hierarchical format for description of graphics representables that can be augmented by physics information and relational properties. It was developed for high energy physics event display applications and is especially suited to client/server or component frameworks. The GLAST experiment, an international effort led by NASA for a gamma-ray telescope to launch in 2006, chose HepRep to provide a flexible, extensible and maintainable framework for their event display without tying their users to any one graphics application. To support HepRep in their GUADI infrastructure, GLAST developed a HepRep filler and builder architecture. The architecture hides the details of XML and CORBA in a set of base and helper classes allowing physics experts to focus on what data they want to represent. GLAST has two GAUDI services: HepRepSvc, which registers HepRep fillers in a global registry and allows the HepRep to be exported to XML, and CorbaSvc, which allows the HepRep to be published through a CORBA interface and which allows the client application to feed commands back to GAUDI (such as start next event, or run some GAUDI algorithm). GLAST s HepRep solution gives users a choice of client applications, WIRED (written in Java) or FRED (written in C++ and Ruby), and leaves them free to move to any future HepRep-compliant event display. HepRep is a generic, hierarchical format for description of graphics representables that can be augmented by physics information and relational properties. It was developed for high energy physics event display applications and is especially suited to client/server or component frameworks. The GLAST experiment, an international effort led by NASA for a gamma-ray telescope to launch in 2006, chose HepRep to provide a flexible, extensible and maintainable framework for their event display without tying their users to any one graphics application. This paper describes why GLAST selected HepRep and how they went about implementing a HepRep-based event display in their GAUDI framework.

A novel background reduction strategy for high level triggers and processing in gamma‐ray Cherenkov detectors

Gamma ray astronomy is now at the leading edge for studies related both to fundamental physics and astrophysics. The sensitivity of gamma detectors is limited by the huge amount of background, constituted by hadronic cosmic rays (typically two to three orders of magnitude more than the signal) and by the accidental background in the detectors. By using the information on the temporal evolution of the Cherenkov light, the background can be reduced. We will present here the results obtained within the MAGIC experiment using a new technique for the reduction of the background. Particle showers produced by gamma rays show a different temporal distribution with respect to showers produced by hadrons; the background due to accidental counts shows no dependence on time. Such novel strategy can increase the sensitivity of present instruments.

A THIRD LEVEL TRIGGER PROGRAMMABLE ON FPGA FOR THE GAMMA/HADRON SEPARATION IN A CHERENKOV TELESCOPE USING PSEUDO-ZERNIKE MOMENTS AND THE SVM CLASSIFIER

We studied the application of the Pseudo-Zernike features as image parameters (instead of the Hillas parameters) for the discrimination between the images produced by atmospheric electromagnetic showers caused by gamma-rays and the ones produced by atmospheric electromagnetic showers caused by hadrons in the MAGIC Experiment. We used a Support Vector Machine as classification algorithm with the computed Pseudo-Zernike features as classification parameters. We implemented on a FPGA board a kernel function of the SVM and the Pseudo-Zernike features to build a third level trigger for the gamma-hadron separation task of the MAGIC Experiment.

SIMULATING THE HIGH ENERGY GAMMA-RAY SKY SEEN BY THE GLAST LARGE AREA TELESCOPE

This paper presents the simulation of the GLAST high energy gamma-ray telescope. The simulation package, written in C++, is based on the Geant4 toolkit, and it is integrated into a general framework used to process events. A detailed simulation of the electronic signals inside Silicon detectors has been provided and it is used for the particle tracking, which is handled by a dedicated software. A unique repository for the geometrical description of the detector has been realized using the XML language and a C++ library to access this information has been designed and implemented. A new event display based on the HepRep protocol was implemented. The full simulation

DATA MINING IN GAMMA ASTROPHYSICS EXPERIMENTS

Data mining techniques, including clustering and classification tasks, for the automatic information extraction from large datasets are increasingly demanded in several scientific fields. In particular, in the astrophysical field, large archives and digital sky surveys with dimensions of 10E12 bytes currently exist, while in the near future they will reach sizes of the order of 10E15. In this work we propose a multidimensional indexing method to efficiently query and mine large astrophysical datasets. A novelty detection algorithm, based on the Support Vector Clustering and using density and neighborhood information stored in the index structure, is proposed to find regions of interest in data characterized by isotropic noise. We show an application of this method for the detection of point sources from a gamma-ray photon list.

A data-driven multidimensional indexing method for datamining in astrophysical databases

Large archives and digital sky surveys with dimensions of bytes currently exist, while in the near future they will reach sizes of the order of . Numerical simulations are also producing comparable volumes of information. Data mining tools are needed for information extraction from such large datasets. In this work, we propose a multidimensional indexing method, based on a static R-tree data structure, to efficiently query and mine large astrophysical datasets. We follow a top-down construction method, called VAMSplit, which recursively splits the dataset on a near median element along the dimension with maximum variance. The obtained index partitions the dataset into nonoverlapping bounding boxes, with volumes proportional to the local data density. Finally, we show an application of this method for the detection of point sources from a gamma-ray photon list.

Perspectives in astrophysical databases

Astrophysics has become a domain extremely rich of scientific data. Data mining tools are needed for information extraction from such large data sets. This asks for an approach to data management emphasizing the efficiency and simplicity of data access; efficiency is obtained using multidimensional access methods and simplicity is achieved by properly handling metadata. Moreover, clustering and classification techniques on large data sets pose additional requirements in terms of computation and memory scalability and interpretability of results. In this study we review some possible solutions.