Nuclear Instruments and Methods in Physics Research A 466 (2001) 538–550 Investigation of t % t in the full hadronic final state at CDF with a neural network approach $ A. Sidoti a, *, P. Azzi b , G. Busetto b , A. Castro b , S. Dusini b , I. Lazzizzera a , J.L. Wyss b a Dipartimento di Fisica, Universit " a di Trento and INFN, Gruppo Collegato Sezione di Padova, Via Sommarive 14, I-38050 Povo, Trento, Italy b Dipartimento di Fisica, Universit " a di Padova and INFN, Sezione di Padova, I-35131 Padova, Italy Received 8 November 2000; accepted 18 January 2001 Abstract In this work we present the results of a neural network (NN) approach to the measurement of the t % t production cross- section and top mass in the all-hadronic channel, analyzing data collected at the Collider Detector at Fermilab (CDF) experiment. We have used a hardware implementation of a feed forward neural network, TOTEM, the product of a collaboration of INFN (Istituto Nazionale Fisica Nucleare)FIRST (Istituto per la Ricerca Scientifica e Tecnologica)FUniversity of Trento, Italy. Particular attention has been payed to the evaluation of the systematics specifically related to the NN approach. The results are consistent with those obtained at CDF by conventional data selection techniques. r 2001 Elsevier Science B.V. All rights reserved. PACS: 84.35; 14.65Ha; 07.05Mh Keywords: Neural network; Hardware; Data analysis; Methodologies; Top quark; Hadronic decay 0. Introduction Neural nets have received growing attention for the last few years, in High Energy Physics (HEP) experiments. Their pattern recognition ability can be used in data analysis to classify events and are particularly useful in situations when the data have probabilistic nature or incomplete characteriza- tion. Using highly-parallel hardware-implemented NN, one can couple high clock rates to quality performance, in order to get on-line operativity, that might be exploited for trigger designs in the next HEP collider experiments (CDFII, ATLAS, CMS,y). The aim of this work is to measure the t % t production cross-section and the top mass in the all-hadronic channel at CDF using the TOTEM neurochip, the product of a joint enterprise, mainly by Istituto Nazionale Fisica Nucleare (INFN). Indeed, apart from the measurement of the physical quantities, the main topics will be the definition of suitable methodologies and quantita- tive evaluation of the systematics specific to neural networks (NN). We recall that the analysis of the $ Partially supported by NIS. *Corresponding author. E-mail address: sidoti@fnal.gov (A. Sidoti). 0168-9002/01/$-see front matter r 2001 Elsevier Science B.V. All rights reserved. PII:S0168-9002(01)00619-2