Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Powiadomienia systemowe

Sesja wygasła!

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: multivariate methods

Sortuj według:

Ogranicz wyniki do:

Optimization of tau indentification in ATLAS experiment using multivariate tools

Janyst Ł., Kaczmarska A., Szymocha T., Wolter M., Zemła A.

Computer Science

2008

Vol. 9

35-45

Elementary particle physics experiments, which search for very rare processes, require the efficient analysis and selection algorithms able to separate a signal from the overwhelming background. Four learning machine algorithms have been applied to identify τ leptons in the ATLAS experiment: projective likelihood estimator (LL), Probability Density Estimator with Range Searches (PDE-RS), Neural Network, and the Support Vector Machine (SVM). All four methods have similar performance, which is significantly better than the baseline cut analysis. This indicates that the achieved background rejection is close to the maximal achievable performance.

W eksperymentach fizyki wysokich energii, poszukujących bardzo rzadkich procesów, dużego znaczenia nabierają algorytmy umożliwiające separację sygnału od przeważającego tła. Cztery algorytmy uczące się na przykładach zostały zastosowane do identyfikacji leptonów tau w eksperymencie ATLAS: rzutowane rozkłady prawdopodobieństw (projective likelihood estimator - LL), PDE-RS (Probability Density Estimator with Range Searches), sieć neuronowa oraz maszyna wektorów wspierających (SVM). Algorytmy te mają zbliżone wydajności znacząco lepsze od standardowej analizy z użyciem cięć. Sugeruje to, że osiągnięte wydajności są bliskie maksymalnej osiągalnej granicy.

Testing the suitability of leeches (Hirudinea, Clitellata) for biological assessment of lowland streams

Koperski P.

Polish Journal of Ecology

2005

Vol. 53, nr 1

65-80

Certain features of leeches make them potentially very useful in the biological assessment of freshwaters as bio-indicators of water pollution, especially in moderately polluted lowland watercourses. The main aim of the study was to test their usefulness as indicators of the level of habitat degradation. The composition of leech samples and the main abiotic parameters in water samples, which had been taken simultaneously in six lowland streams of north-eastern Poland and the relations between them were investigated. Multivariate methods of data analysis wereused to test if particular leech species or the taxonomic composition of assemblages thereof could be useful in predicting the assessed quality of the environment. The greatest difficulty with analysis of the results is to separate the effect of the stream (site specifics) from the effect of environmental quality, expressed as Cumulative Index of Environmental Quality (CIEQ). An important result of this study is the highly consistent prediction of the level of water pollution to be attained on the basis of leech taxonomic composition using Discriminant Function Analysis. Some common leech species were found to be good, positive or negative indicators of pollution level. The classifications of Glossiphonia complanata and Erpobdella octoculata as negative bio-indicators corroborates the conclusions of numerous previous studies, in contrast with the indicative value of Helobdella stagnalis, shown previously to be rather a negative indicator. It would seem to be necessary, for biomonitoring purposes, to identify detailly the leeches of such ecologically different families as Glossiphonidae and Erpobdellidae.