Unbiased sensitivity analysis and pruning techniques in neural networks for surface ozone modelling

2005 ◽  
Vol 182 (2) ◽  
pp. 149-158 ◽  
Author(s):  
O. Pastor-Bárcenas ◽  
E. Soria-Olivas ◽  
J.D. Martín-Guerrero ◽  
G. Camps-Valls ◽  
J.L. Carrasco-Rodríguez ◽  
...  
Keyword(s):  
Neural Networks ◽  
Sensitivity Analysis ◽  
Surface Ozone ◽  
Ozone Modelling ◽  
Pruning Techniques

scholarly journals Pruning algorithms of neural networks — a comparative study

2013 ◽  
Vol 3 (3) ◽  
Author(s):  
M. Augasta ◽  
T. Kathirvalavakumar
Keyword(s):  
Breast Cancer ◽  
Neural Network ◽  
Neural Networks ◽  
Sensitivity Analysis ◽  
Mutual Information ◽  
Learning Process ◽  
The Neural Network ◽  
The Neural Networks ◽  
Pruning Algorithms ◽  
Pruning Techniques

AbstractThe neural network with optimal architecture speeds up the learning process and generalizes the problem well for further knowledge extraction. As a result researchers have developed various techniques for pruning the neural networks. This paper provides a survey of existing pruning techniques that optimize the architecture of neural networks and discusses their advantages and limitations. Also the paper evaluates the effectiveness of various pruning techniques by comparing the performance of some traditional and recent pruning algorithms based on sensitivity analysis, mutual information and significance on four real datasets namely Iris, Wisconsin breast cancer, Hepatitis Domain and Pima Indian Diabetes.

scholarly journals Sensitivity analysis for interpretation of machine learning based segmentation models in cardiac MRI

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Markus J. Ankenbrand ◽  
Liliia Shainberg ◽  
Michael Hock ◽  
David Lohr ◽  
Laura M. Schreiber
Keyword(s):  
Neural Networks ◽  
Sensitivity Analysis ◽  
Cardiac Magnetic Resonance ◽  
Magnetic Resonance ◽  
Cardiac Mri ◽  
Magnetic Resonance Images ◽  
Model Interpretation ◽  
First Case ◽  
Segmentation Models

Abstract Background Image segmentation is a common task in medical imaging e.g., for volumetry analysis in cardiac MRI. Artificial neural networks are used to automate this task with performance similar to manual operators. However, this performance is only achieved in the narrow tasks networks are trained on. Performance drops dramatically when data characteristics differ from the training set properties. Moreover, neural networks are commonly considered black boxes, because it is hard to understand how they make decisions and why they fail. Therefore, it is also hard to predict whether they will generalize and work well with new data. Here we present a generic method for segmentation model interpretation. Sensitivity analysis is an approach where model input is modified in a controlled manner and the effect of these modifications on the model output is evaluated. This method yields insights into the sensitivity of the model to these alterations and therefore to the importance of certain features on segmentation performance. Results We present an open-source Python library (misas), that facilitates the use of sensitivity analysis with arbitrary data and models. We show that this method is a suitable approach to answer practical questions regarding use and functionality of segmentation models. We demonstrate this in two case studies on cardiac magnetic resonance imaging. The first case study explores the suitability of a published network for use on a public dataset the network has not been trained on. The second case study demonstrates how sensitivity analysis can be used to evaluate the robustness of a newly trained model. Conclusions Sensitivity analysis is a useful tool for deep learning developers as well as users such as clinicians. It extends their toolbox, enabling and improving interpretability of segmentation models. Enhancing our understanding of neural networks through sensitivity analysis also assists in decision making. Although demonstrated only on cardiac magnetic resonance images this approach and software are much more broadly applicable.

scholarly journals A State-Based Sensitivity Analysis for Distinguishing the Global Importance of Predictor Variables in Artificial Neural Networks

2016 ◽  
Vol 2016 ◽  
pp. 1-11 ◽  
Author(s):  
Ehsan Ardjmand ◽  
David F. Millie ◽  
Iman Ghalehkhondabi ◽  
William A. Young II ◽  
Gary R. Weckman
Keyword(s):  
Neural Networks ◽  
Sensitivity Analysis ◽  
Artificial Neural Networks ◽  
Network Models ◽  
Predictor Variables ◽  
Apparent Lack ◽  
Artificial Neural ◽  
High Degree ◽  
Universal Approximators ◽  
Empirical Approaches

Artificial neural networks (ANNs) are powerful empirical approaches used to model databases with a high degree of accuracy. Despite their recognition as universal approximators, many practitioners are skeptical about adopting their routine usage due to lack of model transparency. To improve the clarity of model prediction and correct the apparent lack of comprehension, researchers have utilized a variety of methodologies to extract the underlying variable relationships within ANNs, such as sensitivity analysis (SA). The theoretical basis of local SA (that predictors are independent and inputs other than variable of interest remain “fixed” at predefined values) is challenged in global SA, where, in addition to altering the attribute of interest, the remaining predictors are varied concurrently across their respective ranges. Here, a regression-based global methodology, state-based sensitivity analysis (SBSA), is proposed for measuring the importance of predictor variables upon a modeled response within ANNs. SBSA was applied to network models of a synthetic database having a defined structure and exhibiting multicollinearity. SBSA achieved the most accurate portrayal of predictor-response relationships (compared to local SA and Connected Weights Analysis), closely approximating the actual variability of the modeled system. From this, it is anticipated that skepticisms concerning the delineation of predictor influences and their uncertainty domains upon a modeled output within ANNs will be curtailed.

Input layer optimization of neural networks by sensitivity analysis and its application to recognition of numerals

1991 ◽  
Vol 111 (4) ◽  
pp. 130-138 ◽  
Author(s):  
Hiroshi Takenaga ◽  
Shigeo Abe ◽  
Masao Takato ◽  
Masahiro Kayama ◽  
Tadaaki Kitamura ◽  
...  
Keyword(s):  
Neural Networks ◽  
Sensitivity Analysis ◽  
Input Layer

scholarly journals Sensitivity analysis of energy inputs in crop production using artificial neural networks

2018 ◽  
Vol 197 ◽  
pp. 992-998 ◽  
Author(s):  
Alireza Khoshroo ◽  
Ali Emrouznejad ◽  
Ahmadreza Ghaffarizadeh ◽  
Mehdi Kasraei ◽  
Mahmoud Omid
Keyword(s):  
Neural Networks ◽  
Sensitivity Analysis ◽  
Artificial Neural Networks ◽  
Crop Production ◽  
Energy Inputs ◽  
Artificial Neural

scholarly journals Are We Correctly Using Discharge Coefficients for Side Weirs? Insights from a Numerical Investigation

Water ◽  
2019 ◽  
Vol 11 (12) ◽  
pp. 2585
Author(s):  
Mario Di Bacco ◽  
Anna Rita Scorzini
Keyword(s):  
Neural Networks ◽  
Sensitivity Analysis ◽  
Discharge Coefficient ◽  
Experimental Methodology ◽  
Laboratory Measurements ◽  
Experimental Assessment ◽  
Discharge Coefficients ◽  
Side Weirs ◽  
Water Depths

A key issue in the design of side weirs is the experimental assessment of the discharge coefficient. This can be determined by laboratory measurements of discharge and water depths at the up- and downstream ends of the weir by using De Marchi’s approach, consisting in the solution of the 1D dynamic equation of spatially varied steady flow with non-uniform discharge, under the assumption of energy conservation. This study originates from a recent alarming proliferation of works that evaluate the discharge coefficient for side weirs without clearly explaining the experimental methodology and/or even incorrectly applying modelling approaches, thus generating possible misinterpretations of the results. In this context, the present paper aims to highlight the effects of using oversimplified and/or heterogenous models (relying on different assumptions) for the experimental determination of the discharge coefficient for side weirs. Furthermore, a sensitivity analysis is performed to detect the most influencing hydraulic and geometric parameters on each considered model. The overall results clearly indicate the wrongness of using or building not homogeneous discharge coefficient datasets to obtain and/or compare predictive experimental discharge coefficient formulas. We finally show how neural networks could provide a possible solution to these heterogeneity issues.

Towards a Methodology for Monitoring and Analyzing the Supply Chain Behavior

2011 ◽  
pp. 186-208
Author(s):  
Reinaldo Moraga ◽  
Luis Rabelo ◽  
Alfonso Sarmiento
Keyword(s):  
Neural Networks ◽  
Sensitivity Analysis ◽  
Supply Chain ◽  
System Dynamics ◽  
Eigenvalue Analysis ◽  
Background Information ◽  
Short And Long Term ◽  
And Performance

In this chapter, the authors present general steps towards a methodology that contributes to the advancement of prediction and mitigation of undesirable supply chain behavior within short- and long- term horizons by promoting a better understanding of the structure that determines the behavior modes. Through the integration of tools such as system dynamics, neural networks, eigenvalue analysis, and sensitivity analysis, this methodology (1) captures the dynamics of the supply chain, (2) detects changes and predicts the behavior based on these changes, and (3) defines needed modifications to mitigate the unwanted behaviors and performance. In the following sections, some background information is given from literature, the general steps of the proposed methodology are discussed, and finally a case study is briefly summarized.

Sign in / Sign up

Export Citation Format

Share Document