TY - GEN
T1 - Improving Stroke Trace Classification Explainability Through Counterexamples
AU - Leonardi, Giorgio
AU - Montani, Stefania
AU - Striani, Manuel
N1 - Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2023
Y1 - 2023
N2 - Deep learning process trace classification is proving powerful in several application domains, including medical ones; however, classification results are typically not explainable, an issue which is particularly relevant in medicine. In our recent work we tackled this problem, by proposing trace saliency maps, a novel tool able to highlight what trace activities are particularly significant for the classification task. A trace saliency map is built by generating artificial perturbations of the trace at hand that are classified in the same class as the original one, called examples. In this paper, we investigate the role of counterexamples (i.e., artificial perturbations that are classified in a different class with respect to the original trace) in refining trace saliency map information, thus improving explainability. We test the approach in the domain of stroke.
AB - Deep learning process trace classification is proving powerful in several application domains, including medical ones; however, classification results are typically not explainable, an issue which is particularly relevant in medicine. In our recent work we tackled this problem, by proposing trace saliency maps, a novel tool able to highlight what trace activities are particularly significant for the classification task. A trace saliency map is built by generating artificial perturbations of the trace at hand that are classified in the same class as the original one, called examples. In this paper, we investigate the role of counterexamples (i.e., artificial perturbations that are classified in a different class with respect to the original trace) in refining trace saliency map information, thus improving explainability. We test the approach in the domain of stroke.
UR - http://www.scopus.com/inward/record.url?scp=85164001192&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-34344-5_16
DO - 10.1007/978-3-031-34344-5_16
M3 - Conference contribution
AN - SCOPUS:85164001192
SN - 9783031343438
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 125
EP - 129
BT - Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings
A2 - Juarez, Jose M.
A2 - Marcos, Mar
A2 - Stiglic, Gregor
A2 - Tucker, Allan
PB - Springer Science and Business Media Deutschland GmbH
T2 - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023
Y2 - 12 June 2023 through 15 June 2023
ER -