Predoc-seminarium: Zhendong Wang
Seminarium
Datum: tisdag 18 juni 2024
Tid: 13.00 – 15.00
Plats: L30, DSV, Borgarfjordsgatan 12, Kista
Välkommen till ett predoc-seminarium om tolkningsbarhet och tillit när maskininlärning används inom sjukvården. Zhendong Wang, doktorand på DSV, är respondent.
18 juni 2024 presenterar doktoranden Zhendong Wang sitt pågående arbete med titeln ”Constrained Counterfactual Explanations for Temporal Data”. Seminariet genomförs på Institutionen för data- och systemvetenskap (DSV) vid Stockholms universitet.
Respondent: Zhendong Wang, DSV
Opponent: Pedro Pereira Rodrigues, University of Porto, Portugal
Huvudhandledare: Panagiotis Papapetrou, DSV
Handledare: Isak Samsten, DSV
Närmast berörda professor: Rahim Rahmani, DSV
Kontaktuppgifter till Zhendong Wang
Sammanfattning på engelska
In machine learning, achieving high performance in predictive models while maintaining interpretability is crucial in critical domains, such as healthcare. Recent advances in algorithms in handling temporal data have gained success in time series prediction and event sequence modelling.
However, these algorithms remain opaque, which causes difficulty in understanding the prediction outcome and further leads to users not trusting the model. Counterfactual explanations, as a post-hoc explanation method, can provide actionable insights into machine learning models already deployed in different domains.
Specifically, counterfactual explanations can provide intuitive, sample-based explanations by suggesting modifications of the input sample to achieve the desired prediction outcome. There has not been much research on providing counterfactual explanations for temporal data predictions; hence, the thesis focuses on the counterfactual generation algorithms for time series classification, time series forecasting and event sequence classification. Furthermore, applying counterfactual generation requires domain knowledge to guide the counterfactuals to become more relevant.
Previous work has investigated plausible counterfactual changes but has also followed causal graph models to produce the counterfactuals. This thesis investigates applying two constraint mechanisms, local temporal and domain-specific constraints, to provide more relevant and efficient counterfactual explanations.
Senast uppdaterad: 20 maj 2024
Sidansvarig: Institutionen för data- och systemvetenskap, DSV