Interpreting the predictions of neural network classification based on a Taylor Coefficient Analysis (TCA)
Abstract
We introduce a rigid and comprehensive taxonomy and paradigm for characterizing the influence of the input feature space X on the predictions ŷ of a neural network (NN) used for event classification, based on a Taylor expansion of ŷ in X. The complete process of introspection we refer to as Taylor Coefficient Analysis (TCA). Based on two simplistic example tasks, which can be easily understood and bencmarked, we illustrate the power of the TCA when it comes to revealing, what properties of X have led to what value of ŷ, of a given NN model, building up intuition for the method. A more complex application is meant to represent X of a typical classification task at a CERN LHC experiment. Based on this application, we play through the different levels of introspection that the TCA offers and discuss a number of practical aspects for a TCA application typical for the analysis of CERN LHC data. We conclude with a study to support the assumption that those properties of X most relevant for tasks of the complexity typical for a CERN LHC experiment, are usually caught by a TCA up to the second order.
Related
- Neural network-based multipartite entanglement classification with prior guidance from quantum uncertainty relations
- Identifiability Limits of Gravitational Wave Phase Deviations: Multiclass Classification with a Multihead Neural Network
- Search for Diffuse Supernova Neutrino Background in the Full KamLAND Dataset with Neural-Network-Based Event Classification
- Identifying the Shear Piezoelectric Coefficient using Physics-Informed Neural Networks under Exact Boundary and Initial Constraints
- Approximating velocity fields with planted point attractors via Neural-ODE for classification purposes
BibTeX
@misc{klute2026interpreting,
title = {{Interpreting the predictions of neural network classification based on a Taylor Coefficient Analysis (TCA)}},
author = {Markus Klute and Artur Monsch and Lars Sowa and Roger Wolf},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.12801},
eprint = {2609.12801},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.12801},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Interpreting the predictions of neural network classification based on a Taylor Coefficient Analysis (TCA)
AU - Markus Klute
AU - Artur Monsch
AU - Lars Sowa
AU - Roger Wolf
PY - 2026
DA - 2026/09/11
PB - arXiv
DO - 10.48550/arxiv.2609.12801
UR - https://arxiv.org/abs/2609.12801
AB - We introduce a rigid and comprehensive taxonomy and paradigm for characterizing the influence of the input feature space X on the predictions ŷ of a neural network (NN) used for event classification, based on a Taylor expansion of ŷ in X. The complete process of introspection we refer to as Taylor Coefficient Analysis (TCA). Based on two simplistic example tasks, which can be easily understood and bencmarked, we illustrate the power of the TCA when it comes to revealing, what properties of X have led to what value of ŷ, of a given NN model, building up intuition for the method. A more complex application is meant to represent X of a typical classification task at a CERN LHC experiment. Based on this application, we play through the different levels of introspection that the TCA offers and discuss a number of practical aspects for a TCA application typical for the analysis of CERN LHC data. We conclude with a study to support the assumption that those properties of X most relevant for tasks of the complexity typical for a CERN LHC experiment, are usually caught by a TCA up to the second order.
ER -