Layerwise Relevance Visualization in Convolutional Text Graph Classifiers

Schwarzenberg, Robert; Hübner, Marc P.; Harbecke, David; Alt, Christoph; Hennig, Leonhard

doi:10.18653/v1/d19-5308

Cited by 52 publications

(20 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These post-hoc methods are typically used to analyze individual feature input and output pairs, limiting their explainability to an individual-level. Several explanation methods have been presented in the literature including layer-wise relevance propagation (LRP) [ 225 ], excitation backpropagation [ 224 ], graph pruning (GNNExplainer) [ 223 ], gradient-based saliency (GraphGrad-CAM) [ 226 ], GraphGrad-CAM++ [ 227 ]), and layerwise relevance propagation (GraphLRP) [ 228 ].…”

Section: Research Challenges and Future Directionsmentioning

confidence: 99%

Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future

Ahmedt-Aristizabal

Armin

Denman

et al. 2021

Sensors

136

View full text Add to dashboard Cite

With the advances of data-driven machine learning research, a wide variety of prediction problems have been tackled. It has become critical to explore how machine learning and specifically deep learning methods can be exploited to analyse healthcare data. A major limitation of existing methods has been the focus on grid-like data; however, the structure of physiological recordings are often irregular and unordered, which makes it difficult to conceptualise them as a matrix. As such, graph neural networks have attracted significant attention by exploiting implicit information that resides in a biological system, with interacting nodes connected by edges whose weights can be determined by either temporal associations or anatomical junctions. In this survey, we thoroughly review the different types of graph architectures and their applications in healthcare. We provide an overview of these methods in a systematic manner, organized by their domain of application including functional connectivity, anatomical structure, and electrical-based analysis. We also outline the limitations of existing techniques and discuss potential directions for future research.

show abstract

Section: Research Challenges and Future Directionsmentioning

confidence: 99%

Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future

Ahmedt-Aristizabal

Armin

Denman

et al. 2021

Sensors

136

View full text Add to dashboard Cite

show abstract

“…A pioneering work on explanation techniques for GNNs was published in 2015 [251]. In the time since, several explanation methods have been presented including layer-wise relevance propagation (LRP) [252], excitation backpropagation [253], graph pruning (GNNEx-plainer) [250], gradient-based saliency (GRAPHGRAD-CAM) [254], GRAPHGRAD-CAM++ [255]), and layerwise relevance propagation (GRAPHLRP) [256]. Attention mechanisms adopted in several medical applications discussed in our survey have also been used as another explanation technique where the attention weights for edges can be used to measure edge importance; however, it is noted that they can only explain GAT models without explaining node features, unlike, for example, GNNExplainer.…”

Section: B Challenges In Adapting Graph-based Deep Learning Methods F...mentioning

confidence: 99%

Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future

Ahmedt-Aristizabal,

Armin,

Denman

et al. 2021

Preprint

View full text Add to dashboard Cite

With the advances of data-driven machine learning research, a wide variety of prediction problems have been tackled. It has become critical to explore how machine learning and specifically deep learning methods can be exploited to analyse healthcare data. A major limitation of existing methods has been the focus on grid-like data; however, the structure of physiological recordings are often irregular and unordered which makes it difficult to conceptualise them as a matrix. As such, graph neural networks have attracted significant attention by exploiting implicit information that resides in a biological system, with interactive nodes connected by edges whose weights can be either temporal associations or anatomical junctions. In this survey, we thoroughly review the different types of graph architectures and their applications in healthcare. We provide an overview of these methods in a systematic manner, organized by their domain of application including functional connectivity, anatomical structure and electrical-based analysis. We also outline the limitations of existing techniques and discuss potential directions for future research.

show abstract

“…The work [24] extends explanation techniques such as Grad-CAM or Excitation Backprop to the GNN model, and arrives at an attribution on nodes of the graph. In an NLP context, graph convolutional networks (GCNs) have been explained in terms of nodes and edges in the input graph using the LRP explanation method [25]. GNNExplainer [26] and PGExplainer [27] explain the model by extracting the subgraph that maximizes the mutual information to the prediction for the original graph.…”

Section: Explaining Graph Neural Networkmentioning

confidence: 99%

Higher-Order Explanations of Graph Neural Networks via Relevant Walks

Schnake

Eberle

Lederer

et al. 2022

IEEE Trans. Pattern Anal. Mach. Intell.

114

View full text Add to dashboard Cite

Graph Neural Networks (GNNs) are a popular approach for predicting graph structured data. As GNNs tightly entangle the input graph into the neural network structure, common explainable AI approaches are not applicable. To a large extent, GNNs have remained black-boxes for the user so far. In this paper, we show that GNNs can in fact be naturally explained using higher-order expansions, i.e. by identifying groups of edges that jointly contribute to the prediction. Practically, we find that such explanations can be extracted using a nested attribution scheme, where existing techniques such as layer-wise relevance propagation (LRP) can be applied at each step. The output is a collection of walks into the input graph that are relevant for the prediction. Our novel explanation method, which we denote by GNN-LRP, is applicable to a broad range of graph neural networks and lets us extract practically relevant insights on sentiment analysis of text data, structure-property relationships in quantum chemistry, and image classification.Index Terms-graph neural networks, higher-order explanations, layer-wise relevance propagation, explainable machine learning. ! INTRODUCTIONMany interesting structures found in scientific and industrial applications can be expressed as graphs. Examples are lattices in fluid modeling, molecular geometry, biological interaction networks, or social / historical networks. Graph neural networks (GNNs) [1], [2] have been proposed as a method to learn from observations in general graph structures and have found use in an ever growing number of applications [3]-[8]. While GNNs make useful predictions, they typically act as black-boxes, and it has neither been directly possible (1) to extract novel insight from the learned model nor (2) to verify that the model has made the intended use of the graph structure, e.g. that it has avoided Clever Hans phenomena [9].Explainable AI (XAI) is an emerging research area that aims to extract interpretable insights from trained ML models [10], [11]. So far, research has focused, for example, on full black-box models [12], [13], self-explainable models [14], [15], or deep neural networks [16], where in all cases, the prediction can be attributed to the input features. For a GNN, however, the graph being received as input is deeply

show abstract

Layerwise Relevance Visualization in Convolutional Text Graph Classifiers

Cited by 52 publications

References 12 publications

Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future

Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future

Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future

Higher-Order Explanations of Graph Neural Networks via Relevant Walks

Contact Info

Product

Resources

About