Do We Need Another Explainable AI Method? Toward Unifying Post-hoc XAI Evaluation Methods into an Interactive and Multi-dimensional Benchmark

Belaid, Mohamed Karim; Hüllermeier, Eyke; Rabus, Maximilian; Krestel, Ralf

doi:10.48550/arxiv.2207.14160

Cited by 5 publications

(5 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For such comparisons, XAI Toolsheets adopt 22 dimensions falling under three major categories, namely metadata, utility, and usability. Aside from documentations, Belaid et al [192] introduce Compare-xAI, a unified benchmark, with multiple use-cases, indexing +16 post-hoc xAI algorithms, +22 tests, and +40 research paper. Through Compare-xAI, practitioners and data scientists can gain insights on which XAI methods are relevant to their problems.…”

Section: Xai Methods Selectionmentioning

confidence: 99%

XAIR: A Systematic Metareview of Explainable AI (XAI) Aligned to the Software Development Process

Clement¹,

Kemmerzell²,

Abdelaal

et al. 2023

MAKE

View full text Add to dashboard Cite

Currently, explainability represents a major barrier that Artificial Intelligence (AI) is facing in regard to its practical implementation in various application domains. To combat the lack of understanding of AI-based systems, Explainable AI (XAI) aims to make black-box AI models more transparent and comprehensible for humans. Fortunately, plenty of XAI methods have been introduced to tackle the explainability problem from different perspectives. However, due to the vast search space, it is challenging for ML practitioners and data scientists to start with the development of XAI software and to optimally select the most suitable XAI methods. To tackle this challenge, we introduce XAIR, a novel systematic metareview of the most promising XAI methods and tools. XAIR differentiates itself from existing reviews by aligning its results to the five steps of the software development process, including requirement analysis, design, implementation, evaluation, and deployment. Through this mapping, we aim to create a better understanding of the individual steps of developing XAI software and to foster the creation of real-world AI applications that incorporate explainability. Finally, we conclude with highlighting new directions for future research.

show abstract

Section: Xai Methods Selectionmentioning

confidence: 99%

XAIR: A Systematic Metareview of Explainable AI (XAI) Aligned to the Software Development Process

Clement¹,

Kemmerzell²,

Abdelaal

et al. 2023

MAKE

View full text Add to dashboard Cite

show abstract

“…XAI often concerns purely computational approaches but explainability can range from writing algorithms for XAI to publishing and describing a model in an algorithmic registry to stress-testing a model with different data as seen in AI auditing (Costanza-Chock et al, 2022;Gilpin et al, 2018). AI itself can be viewed as a kind of explanation; dimension reduction algorithms (e.g., linear discriminant analysis) can suggest the importance of certain input variables since they can collapse the number of variables while preserving their "essence" (Belaid et al, 2022). This extension of explanation blurs the boundary between the explainability and interpretability within XAI (Gunning et al, 2019).…”

Section: G Eog R Aphi C Appli C Ati On S Of X Ai Me Thods: S Tate-of-...mentioning

confidence: 99%

“…Beyond these challenges, we must ask what constitutes explainability as it spans transparency demands of plain text as a form of explanation to computational remedies for training data, hyperparameter tuning and layer restructuring. We considered how XAI could bypass an explanation of DNN outcomes to directly provide insights about the data (Belaid et al, 2022). We discussed the need for semantic understanding, not only to introduce greater geographic semantics and ontologies into the underlying DNN but also to account for a lack of definitional overlap between the fields of geography and AI.…”

Section: Con Clus Ionmentioning

confidence: 99%

The challenges of integrating explainable artificial intelligence into GeoAI

Jin

Sieber

2023

Transactions in GIS

View full text Add to dashboard Cite

Although explainable artificial intelligence (XAI) promises considerable progress in glassboxing deep learning models, there are challenges in applying XAI to geospatial artificial intelligence (GeoAI), specifically geospatial deep neural networks (DNNs). We summarize these as three major challenges, related generally to XAI computation, to GeoAI and geographic data handling, and to geosocial issues. XAI computation includes the difficulty of selecting reference data/models and the shortcomings of attributing explanatory power to gradients, as well as the difficulty in accommodating geographic scale, geovisualization, and underlying geographic data structures. Geosocial challenges encompass the limitations of knowledge scope—semantics and ontologies—in the explanation of GeoAI as well as the lack of integrating non‐technical aspects in XAI, including processes that are not amenable to XAI. We illustrate these issues with a land use classification case study.

show abstract

“…Another tool for evaluating the XAI explainers is CompareXAI [55] which uses metrics including comprehensibility, portability, and average execution time. Further contribution includes the Local Explanation Evaluation Framework (LEAF) [56] which can evaluate the explanation produced by SHAP and LIME with respect to stability, local concordance, fidelity, and prescriptivity.…”

Section: Related Workmentioning

confidence: 99%

Explainable AI Evaluation: A Top-Down Approach for Selecting Optimal Explanations for Black Box Models

Mirzaei,

Mao,

Al-Nima

et al. 2023

Information

View full text Add to dashboard Cite

Explainable Artificial Intelligence (XAI) evaluation has grown significantly due to its extensive adoption, and the catastrophic consequence of misinterpreting sensitive data, especially in the medical field. However, the multidisciplinary nature of XAI research resulted in diverse scholars possessing significant challenges in designing proper evaluation methods. This paper proposes a novel framework of a three-layered top-down approach on how to arrive at an optimal explainer, accenting the persistent need for consensus in XAI evaluation. This paper also investigates a critical comparative evaluation of explanations in both model agnostic and specific explainers including LIME, SHAP, Anchors, and TabNet, aiming to enhance the adaptability of XAI in a tabular domain. The results demonstrate that TabNet achieved the highest classification recall followed by TabPFN, and XGBoost. Additionally, this paper develops an optimal approach by introducing a novel measure of relative performance loss with emphasis on faithfulness and fidelity of global explanations by quantifying the extent to which a model’s capabilities diminish when eliminating topmost features. This addresses a conspicuous gap in the lack of consensus among researchers regarding how global feature importance impacts classification loss, thereby undermining the trust and correctness of such applications. Finally, a practical use case on medical tabular data is provided to concretely illustrate the findings.

show abstract

Do We Need Another Explainable AI Method? Toward Unifying Post-hoc XAI Evaluation Methods into an Interactive and Multi-dimensional Benchmark

Cited by 5 publications

References 15 publications

XAIR: A Systematic Metareview of Explainable AI (XAI) Aligned to the Software Development Process

XAIR: A Systematic Metareview of Explainable AI (XAI) Aligned to the Software Development Process

The challenges of integrating explainable artificial intelligence into GeoAI

Explainable AI Evaluation: A Top-Down Approach for Selecting Optimal Explanations for Black Box Models

Contact Info

Product

Resources

About