English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Counterfactual Explanations (CEs) are an important tool in AlgorithmicRecourse for addressing two questions: 1. What are the crucial factors that ledto an automated prediction/decision? 2. How can these factors be changed toachieve a more favorable outcome from a user's perspective? Thus, guiding theuser's interaction with AI systems by proposing easy-to-understand explanationsand easy-to-attain feasible changes is essential for the trustworthy adoptionand long-term acceptance of AI systems. In the literature, various methods havebeen proposed to generate CEs, and different quality measures have beensuggested to evaluate these methods. However, the generation of CEs is usuallycomputationally expensive, and the resulting suggestions are unrealistic andthus non-actionable. In this paper, we introduce a new method to generate CEsfor a pre-trained binary classifier by first shaping the latent space of anautoencoder to be a mixture of Gaussian distributions. CEs are then generatedin latent space by linear interpolation between the query sample and thecentroid of the target class. We show that our method maintains thecharacteristics of the input sample during the counterfactual search. Invarious experiments, we show that the proposed method is competitive based ondifferent quality measures on image and tabular datasets -- efficiently returnsresults that are closer to the original data manifold compared to threestate-of-the-art methods, which are essential for realistic high-dimensionalmachine learning applications.

Counterfactual Explanation via Search in Gaussian Mixture Distributed Latent Space