Embedded Feature Importance Determination Technique for Deep Neural Networks- Based Prognostics and Health Management

In the last five years, the inclusion of Deep Learning algorithms in prognostics and health management (PHM) has led to a performance increase in diagnostics, prognostics, and anomaly detection. However, the lack of interpretability of these models results in resistance towards their deployment. Deep Learning-based models fall within the accuracy/interpretability tradeoff, which means that their complexity leads to high performance levels but lacks interpretability. This work aims at addressing this tradeoff by proposing a technique for feature selection embedded in deep neural networks that uses a feature selection (FS) layer trained with the rest of the network to evaluate the input features’ importance. The importance values are used to determine which will be considered for deployment of a PHM model. For comparison with other techniques, this paper introduces a new metric called ranking quality score (RQS), that measures how performance evolves while following the corresponding ranking. The proposed framework is exemplified with three case studies involving health state diagnostics and prognostics and remaining useful life prediction. Results show that the proposed technique achieves higher RQS than the compared techniques, while maintaining the same performance level when compared to the same model but without an FS layer.

Download Full-text

Classifier-Based Evaluation of Image Feature Importance

10.29007/p655 ◽

2018 ◽

Author(s):

Sai Prabhakar Pandi Selvaraj ◽

Manuela Veloso ◽

Stephanie Rosenthal

Keyword(s):

Neural Networks ◽

Image Classification ◽

Deep Neural Networks ◽

Relative Size ◽

Objective Measure ◽

Image Feature ◽

Important Region ◽

Feature Importance ◽

Image Pixels

Significant advances in the performance of deep neural networks, such as Convolutional Neural Networks (CNNs) for image classification, have created a drive for understanding how they work. Different techniques have been proposed to determine which features (e.g., image pixels) are most important for a CNN’s classification. However, the important features output by these techniques have typically been judged subjectively by a human to assess whether the important features capture the features relevant to the classification and not whether the features were actually important to classifier itself. We address the need for an objective measure to assess the quality of different feature importance measures. In particular, we propose measuring the ratio of a CNN’s accuracy on the whole image com- pared to an image containing only the important features. We also consider scaling this ratio by the relative size of the important region in order to measure the conciseness. We demonstrate that our measures correlate well with prior subjective comparisons of important features, but importantly do not require their human studies. We also demonstrate that the features on which multiple techniques agree are important have a higher impact on accuracy than those features that only one technique finds.

Download Full-text

Brain Age Prediction With Morphological Features Using Deep Neural Networks: Results From Predictive Analytic Competition 2019

Frontiers in Psychiatry ◽

10.3389/fpsyt.2020.619629 ◽

2021 ◽

Vol 11 ◽

Author(s):

Angela Lombardi ◽

Alfonso Monaco ◽

Giacinto Donvito ◽

Nicola Amoroso ◽

Roberto Bellotti ◽

...

Keyword(s):

Neural Networks ◽

Deep Neural Networks ◽

Morphological Changes ◽

Structure Characterization ◽

Feature Importance ◽

Mri Scans ◽

Magnetic Resonance Imaging Mri ◽

Brain Age ◽

The Brain ◽

Age Prediction

Morphological changes in the brain over the lifespan have been successfully described by using structural magnetic resonance imaging (MRI) in conjunction with machine learning (ML) algorithms. International challenges and scientific initiatives to share open access imaging datasets also contributed significantly to the advance in brain structure characterization and brain age prediction methods. In this work, we present the results of the predictive model based on deep neural networks (DNN) proposed during the Predictive Analytic Competition 2019 for brain age prediction of 2638 healthy individuals. We used FreeSurfer software to extract some morphological descriptors from the raw MRI scans of the subjects collected from 17 sites. We compared the proposed DNN architecture with other ML algorithms commonly used in the literature (RF, SVR, Lasso). Our results highlight that the DNN models achieved the best performance with MAE = 4.6 on the hold-out test, outperforming the other ML strategies. We also propose a complete ML framework to perform a robust statistical evaluation of feature importance for the clinical interpretability of the results.

Download Full-text

A System-Level Prognostics and Health Management Framework based on Graph Convolutional Neural Networks

Proceedings of the 30th European Safety and Reliability Conference and 15th Probabilistic Safety Assessment and Management Conference ◽

10.3850/978-981-14-8593-0_4165-cd ◽

2020 ◽

Author(s):

Andrés Ruiz-Tagle Palazuelos ◽

Enrique López Droguett ◽

Katrina M. Groth

Keyword(s):

Neural Networks ◽

Convolutional Neural Networks ◽

Health Management ◽

System Level ◽

Prognostics And Health Management ◽

Management Framework

Download Full-text

Early stabilizing feature importance for TensorFlow deep neural networks

2017 International Joint Conference on Neural Networks (IJCNN) ◽

10.1109/ijcnn.2017.7966442 ◽

2017 ◽

Cited By ~ 5

Author(s):

Jeff Heaton ◽

Steven McElwee ◽

James Fraley ◽

James Cannady

Keyword(s):

Neural Networks ◽

Deep Neural Networks ◽

Feature Importance

Download Full-text

Construction and Evaluation of Intelligent Medical Diagnosis Model Based on Integrated Deep Neural Network

Computational Intelligence and Neuroscience ◽

10.1155/2021/7171816 ◽

2021 ◽

Vol 2021 ◽

pp. 1-10

Author(s):

Lina Ma ◽

Tao Yang

Keyword(s):

Neural Networks ◽

Deep Neural Networks ◽

Imaging System ◽

Health Management ◽

Human Life ◽

Diagnostic Model ◽

Quality Healthcare ◽

Model Based ◽

Medical Diagnostic ◽

Demand For Health

In recent years, as human life expectancy increases, birth rate decreases and health management concerns; the traditional Healthcare imaging system, with its uneven Healthcare imaging resources, high Healthcare imaging costs, and diagnoses often relying on doctors’ clinical experience and equipment level limitations, has affected people’s demand for health, so there is a need for a more accurate, convenient, and affordable Healthcare imaging system that allows all people to enjoy fair and quality Healthcare imaging services. This paper discusses the construction and evaluation of an intelligent medical diagnostic model based on integrated deep neural networks, which not only provides a systematic diagnostic analysis of the various symptoms input by the inquirer but also has higher accuracy and efficiency compared with traditional medical diagnostic models. The construction of this model provides a theoretical basis for integrating deep neural networks applied to medical neighborhoods with big data algorithms.

Download Full-text