Matches in SemOpenAlex for { <https://semopenalex.org/work/W4319781094> ?p ?o ?g. }
Showing items 1 to 84 of
84
with 100 items per page.
- W4319781094 abstract "With the rapid development of machine learning, improving its explainability has become a crucial research goal. We study the problem of making the clusters more explainable by investigating the cluster descriptors. Given a set of objects <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$S$</tex> , a clustering of these objects <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$pi$</tex> , and a set of tags <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$T$</tex> that have not participated in the clustering algorithm. Each object in <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$S$</tex> is associated with a subset of <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$T$</tex> . The goal is to find a representative set of tags for each cluster, referred to as the cluster descriptors, with the constraint that these descriptors we find are pairwise disjoint, and the total size of all the descriptors is minimized. In general, this problem is NP-hard. We propose a novel explainability model that reinforces the previous models in such a way that tags that do not contribute to explainability and do not sufficiently distinguish between clusters are not added to the optimal descriptors. The proposed model is formulated as a quadratic unconstrained binary optimization problem which makes it suitable for solving on modern optimization hardware accelerators. We experimentally demonstrate how a proposed explainability model can be solved on specialized hardware for accelerating combinatorial optimization, the Fujitsu Digital Annealer, and use real-life Twitter and PubMed datasets for use cases. Reproducibility materials: Link" @default.
- W4319781094 created "2023-02-11" @default.
- W4319781094 creator A5003048691 @default.
- W4319781094 creator A5032544208 @default.
- W4319781094 creator A5057164472 @default.
- W4319781094 creator A5083306739 @default.
- W4319781094 creator A5090227260 @default.
- W4319781094 date "2022-11-01" @default.
- W4319781094 modified "2023-09-27" @default.
- W4319781094 title "Towards Practical Explainability with Cluster Descriptors" @default.
- W4319781094 cites W2003648292 @default.
- W4319781094 cites W2024860775 @default.
- W4319781094 cites W2039122980 @default.
- W4319781094 cites W2050206309 @default.
- W4319781094 cites W2067612530 @default.
- W4319781094 cites W2083225388 @default.
- W4319781094 cites W2095293504 @default.
- W4319781094 cites W2114853176 @default.
- W4319781094 cites W2159739530 @default.
- W4319781094 cites W2533968486 @default.
- W4319781094 cites W2591456992 @default.
- W4319781094 cites W2887230019 @default.
- W4319781094 cites W2898784085 @default.
- W4319781094 cites W2946426480 @default.
- W4319781094 cites W2948980992 @default.
- W4319781094 cites W2963546261 @default.
- W4319781094 cites W2998562810 @default.
- W4319781094 cites W3093787311 @default.
- W4319781094 cites W3102047970 @default.
- W4319781094 cites W3110558526 @default.
- W4319781094 cites W3127043585 @default.
- W4319781094 cites W3133084444 @default.
- W4319781094 cites W4211250195 @default.
- W4319781094 cites W4234536190 @default.
- W4319781094 doi "https://doi.org/10.1109/icdmw58026.2022.00036" @default.
- W4319781094 hasPublicationYear "2022" @default.
- W4319781094 type Work @default.
- W4319781094 citedByCount "0" @default.
- W4319781094 crossrefType "proceedings-article" @default.
- W4319781094 hasAuthorship W4319781094A5003048691 @default.
- W4319781094 hasAuthorship W4319781094A5032544208 @default.
- W4319781094 hasAuthorship W4319781094A5057164472 @default.
- W4319781094 hasAuthorship W4319781094A5083306739 @default.
- W4319781094 hasAuthorship W4319781094A5090227260 @default.
- W4319781094 hasConcept C114614502 @default.
- W4319781094 hasConcept C119857082 @default.
- W4319781094 hasConcept C124101348 @default.
- W4319781094 hasConcept C154945302 @default.
- W4319781094 hasConcept C177264268 @default.
- W4319781094 hasConcept C184898388 @default.
- W4319781094 hasConcept C199360897 @default.
- W4319781094 hasConcept C23123220 @default.
- W4319781094 hasConcept C33923547 @default.
- W4319781094 hasConcept C41008148 @default.
- W4319781094 hasConcept C45340560 @default.
- W4319781094 hasConcept C73555534 @default.
- W4319781094 hasConceptScore W4319781094C114614502 @default.
- W4319781094 hasConceptScore W4319781094C119857082 @default.
- W4319781094 hasConceptScore W4319781094C124101348 @default.
- W4319781094 hasConceptScore W4319781094C154945302 @default.
- W4319781094 hasConceptScore W4319781094C177264268 @default.
- W4319781094 hasConceptScore W4319781094C184898388 @default.
- W4319781094 hasConceptScore W4319781094C199360897 @default.
- W4319781094 hasConceptScore W4319781094C23123220 @default.
- W4319781094 hasConceptScore W4319781094C33923547 @default.
- W4319781094 hasConceptScore W4319781094C41008148 @default.
- W4319781094 hasConceptScore W4319781094C45340560 @default.
- W4319781094 hasConceptScore W4319781094C73555534 @default.
- W4319781094 hasLocation W43197810941 @default.
- W4319781094 hasOpenAccess W4319781094 @default.
- W4319781094 hasPrimaryLocation W43197810941 @default.
- W4319781094 hasRelatedWork W1608723600 @default.
- W4319781094 hasRelatedWork W197235377 @default.
- W4319781094 hasRelatedWork W2024994743 @default.
- W4319781094 hasRelatedWork W2025362236 @default.
- W4319781094 hasRelatedWork W2370909876 @default.
- W4319781094 hasRelatedWork W2550857612 @default.
- W4319781094 hasRelatedWork W2787810245 @default.
- W4319781094 hasRelatedWork W2975827485 @default.
- W4319781094 hasRelatedWork W3153372913 @default.
- W4319781094 hasRelatedWork W3192150692 @default.
- W4319781094 isParatext "false" @default.
- W4319781094 isRetracted "false" @default.
- W4319781094 workType "article" @default.