A Graph Convolutional Network-Based Approach for Community Detection in Attributed Networks

Wang, Zaisheng; Shen, Guodong; Zhang, Zengjie; Quan, Daying; Li, Jianhua

doi:10.1109/icaice54393.2021.00068

Cited by 2 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Guiding action without causal knowledge is impossible; when causal knowledge is available, our work provides a normative framework for improvement-focused recourse recommendations. Thus, we join a range of work in explainability (Frye, Rowat, and Feige 2020;Heskes et al 2020;Wang, Wiens, and Lundberg 2021;Zhao and Hastie 2021) and fairness (Kilbertus et al 2017;Kusner et al 2017;Zhang and Bareinboim 2018;Makhlouf, Zhioua, and Palamidessi 2020) that highlights the importance of causal knowledge.…”

Section: Limitations and Discussionmentioning

confidence: 99%

Improvement-Focused Causal Recourse (ICR)

König

Freiesleben

Grosse-Wentrup

2023

AAAI

View full text Add to dashboard Cite

Algorithmic recourse recommendations inform stakeholders of how to act to revert unfavorable decisions. However, existing methods may recommend actions that lead to acceptance (i.e., revert the model's decision) but do not lead to improvement (i.e., may not revert the underlying real-world state). To recommend such actions is to recommend fooling the predictor. We introduce a novel method, Improvement-Focused Causal Recourse (ICR), which involves a conceptual shift: Firstly, we require ICR recommendations to guide toward improvement. Secondly, we do not tailor the recommendations to be accepted by a specific predictor. Instead, we leverage causal knowledge to design decision systems that predict accurately pre- and post-recourse, such that improvement guarantees translate into acceptance guarantees. Curiously, optimal pre-recourse classifiers are robust to ICR actions and thus suitable post-recourse. In semi-synthetic experiments, we demonstrate that given correct causal knowledge ICR, in contrast to existing approaches, guides toward both acceptance and improvement.

show abstract

Section: Limitations and Discussionmentioning

confidence: 99%

Improvement-Focused Causal Recourse (ICR)

König

Freiesleben

Grosse-Wentrup

2023

AAAI

View full text Add to dashboard Cite

show abstract

“…The output layer can take several different forms depending on the specific application of the GCN. For example, in the case of ADR detection, the output layer may consist of a binary classification layer that predicts whether a given drug combination is likely to result in an adverse drug reaction, as mentioned in [28]. A neural network has an input, an output and one or many hidden layers.…”

Section: The Need For Predictive Modelsmentioning

confidence: 99%

Evaluating The Effectiveness Of Graph Convolutional Network For Detection Of Polypharmacy Side Effects

IBRAHIM,

Mohammed,

DARA

2023

Preprint

View full text Add to dashboard Cite

Polypharmacy is frequently used to treat numerous illnesses; however, it might have unintended consequences. Recent research using graph convolutional networks (GCNs) has demonstrated promising results in the difficult problem of polypharmacy side effect detection. Using accuracy (ACC), area under the receiver operating characteristic curve (AUC), and F1-score, this study presents a GCN-based model for detecting polypharmacy side effects. The model draws on pharmaceutical data from electronic health records to generate a medication network graph to learn the intricate interrelationships between medications and estimate the risk of adverse effects. The experimental results demonstrate that the proposed GCN-based model outperforms conventional machine learning techniques with an ACC of 91%, an AUC of 0.88, and an F1-score of 0.83. The findings of this article suggest that GCNs are a useful method for monitoring patients on several medications for adverse reactions with total accuracy of 98.47%. Additional study is required to refine the model and collect more data for use in actual clinical settings.

show abstract

A Graph Convolutional Network-Based Approach for Community Detection in Attributed Networks

Cited by 2 publications

References 20 publications

Improvement-Focused Causal Recourse (ICR)

Improvement-Focused Causal Recourse (ICR)

Evaluating The Effectiveness Of Graph Convolutional Network For Detection Of Polypharmacy Side Effects

Contact Info

Product

Resources

About