-
Analysing Neural Network Topologies: a Game Theoretic Approach
Authors:
Julian Stier,
Gabriele Gianini,
Michael Granitzer,
Konstantin Ziegler
Abstract:
Artificial Neural Networks have shown impressive success in very different application cases. Choosing a proper network architecture is a critical decision for a network's success, usually done in a manual manner. As a straightforward strategy, large, mostly fully connected architectures are selected, thereby relying on a good optimization strategy to find proper weights while at the same time avo…
▽ More
Artificial Neural Networks have shown impressive success in very different application cases. Choosing a proper network architecture is a critical decision for a network's success, usually done in a manual manner. As a straightforward strategy, large, mostly fully connected architectures are selected, thereby relying on a good optimization strategy to find proper weights while at the same time avoiding overfitting. However, large parts of the final network are redundant. In the best case, large parts of the network become simply irrelevant for later inferencing. In the worst case, highly parameterized architectures hinder proper optimization and allow the easy creation of adverserial examples fooling the network. A first step in removing irrelevant architectural parts lies in identifying those parts, which requires measuring the contribution of individual components such as neurons. In previous work, heuristics based on using the weight distribution of a neuron as contribution measure have shown some success, but do not provide a proper theoretical understanding. Therefore, in our work we investigate game theoretic measures, namely the Shapley value (SV), in order to separate relevant from irrelevant parts of an artificial neural network. We begin by designing a coalitional game for an artificial neural network, where neurons form coalitions and the average contributions of neurons to coalitions yield to the Shapley value. In order to measure how well the Shapley value measures the contribution of individual neurons, we remove low-contributing neurons and measure its impact on the network performance. In our experiments we show that the Shapley value outperforms other heuristics for measuring the contribution of neurons.
△ Less
Submitted 17 April, 2019;
originally announced April 2019.
-
The upgraded Outer em calorimeter of FOCUS at Fermilab
Authors:
S. Bianco,
F. L. Fabbri,
M. Giardoni,
L. Passamonti,
V. Russo,
S. Sarwar,
A. Zallo,
S. Carrillo,
H. Mendez,
G. Gianini,
J. Anjos,
I. Bediaga,
C. Gobel,
A. Laudo,
J. Magnin,
J. Miranda,
I. Pepe,
F. Simao,
A. Sanchez,
A. Reis
Abstract:
Operational performance, algorithms, stability and physics results of the
Outer em calorimeter of FOCUS are overviewed.
Operational performance, algorithms, stability and physics results of the
Outer em calorimeter of FOCUS are overviewed.
△ Less
Submitted 30 December, 1999;
originally announced December 1999.
-
Cumulant to Factorial Moment Ratio and Multiplicity Data
Authors:
I. M. Dremin,
V. Arena,
G. Boca,
G. Gianini,
S. Malvezzi,
M. Merlo,
S. P. Ratti,
C. Riccardi,
G. Salvadori,
L. Viola,
P. Vitulo
Abstract:
The ratio of cumulant to factorial moments of experimental multiplicity distributions has been calculated for $e^{+}e^{-}$ and $hh$ interactions in a wide range of energies. As a function of the rank it exhibits an initial steep decrease and a series of oscillations around zero. Those features cannot be reproduced by the Negative Binomial Distribution. A comparable behaviour is instead predicted…
▽ More
The ratio of cumulant to factorial moments of experimental multiplicity distributions has been calculated for $e^{+}e^{-}$ and $hh$ interactions in a wide range of energies. As a function of the rank it exhibits an initial steep decrease and a series of oscillations around zero. Those features cannot be reproduced by the Negative Binomial Distribution. A comparable behaviour is instead predicted in high-energy perturbative QCD. The presence of a qualitatively similar behaviour for different processes and in wide energy intervals suggests speaking of an approximate scaling of the cumulant to factorial moment ratio.
△ Less
Submitted 18 May, 1994;
originally announced May 1994.
-
Cumulant to Factorial Moment Ratio and Multiplicity Distributions
Authors:
G. Gianini,
I. M. Dremin,
V. A. Nechitailo,
B. B. Levchenko,
V. Arena,
G. Boca,
S. Malvezzi,
M. Merlo,
S. P. Ratti,
C. Riccardi,
G. Salvadori,
L. Viola,
P. Vitulo
Abstract:
The ratio of cumulant to factorial moments of multiplicity distribu- tions has been calculated for e+e- and hh data in a wide range of energies. As a function of the rank it exhibits a regular behaviour with a steep descent and two negative minima. This behaviour is not accounted for by NBD, that leads to a monotonically decreasing positive-defined ratio. A rough comparison with QCD predictions…
▽ More
The ratio of cumulant to factorial moments of multiplicity distribu- tions has been calculated for e+e- and hh data in a wide range of energies. As a function of the rank it exhibits a regular behaviour with a steep descent and two negative minima. This behaviour is not accounted for by NBD, that leads to a monotonically decreasing positive-defined ratio. A rough comparison with QCD predictions is done: data cannot be accounted for by LLA nor by DLA; only some next-to-next to leading predictions are in qualitative agreement with them. Such features being appropriate for hard processes and high energies, it is a surpri- se to find them not only in e+e- annihilation, but also in soft hh collisions.
△ Less
Submitted 16 December, 1993;
originally announced December 1993.
