Time series classification is a task which deals with temporal sequences, a prevalent data type common in domains such as human activity recognition, sports analytics and general sensing. In this area, interest in explanability has been growing as explanation is key to understand the data and the model better. Recently, a great variety of techniques (e.g., LIME, SHAP, CAM) have been proposed and adapted for time series to provide explanation in the form of , where the importance of each data point in the time series is quantified with a numerical value. However, the saliency maps can and often disagree, so it is unclear which one to use. This paper provides a novel framework to . We show how to robustly evaluate the informativeness of a given explanation method (i.e., relevance for the classification task), and how to compare explanations side-by-side. We propose AMEE, a Model-Agnostic Explanation Evaluation framework, for recommending saliency-based explanations for time series classification. In this approach, data perturbation is added to the input time series guided by each explanation. Our results show that perturbing discriminative parts of the time series leads to significant changes in classification accuracy, which can be used to evaluate each explanation. To be robust to different types of perturbations and different types of classifiers, we aggregate the accuracy loss across perturbations and classifiers. This novel approach allows us to recommend the best explainer among a set of different explainers, including random and oracle explainers. We provide a quantitative and qualitative analysis for synthetic datasets, a variety of time-series datasets, as well as a real-world case study with known expert ground truth.
Download full-text PDF |
Source |
---|---|
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC11513768 | PMC |
http://dx.doi.org/10.1007/s10618-024-01045-8 | DOI Listing |
Int J Integr Care
December 2024
Radboud university medical center, Department of IQ Health, Nijmegen, The Netherlands.
Introduction: Multimorbidity challenges quality and sustainability of healthcare systems. Care groups were introduced in the Netherlands to promote integration of chronic primary care, but it remains unknown to which degree they facilitate this. This study therefore aims to determine whether Dutch general practices perceive themselves to be capable of delivering integrated chronic care and uncover the role of care groups.
View Article and Find Full Text PDFFront Oncol
December 2024
Department of Radiation Oncology, Fudan University Shanghai Cancer Center, Shanghai, China.
Background: For esophageal squamous cell carcinoma (ESCC), universally accepted pathological criteria for classification by differentiation degree are lacking. Tumor budding, single-cell invasion, and nuclear grade, recognized as prognostic factors in other carcinomas, have rarely been investigated for their correlation with differentiation and prognosis in ESCC. This study aims to determine if pathological findings can predict differentiation degree and prognosis in ESCC.
View Article and Find Full Text PDFTher Adv Vaccines Immunother
December 2024
Centre for Human Drug Research, Leiden, The Netherlands.
Respiratory syncytial virus (RSV) causes high worldwide infant mortality, as well as a high disease burden in the elderly. Efforts in vaccine development over the past 60 years have recently delivered three approved vaccines and two monoclonal antibodies (mAbs). Looking back at the eventful history of RSV vaccine development, several factors can be identified that have hampered the developmental pathway, including the occurrence of enhanced RSV disease (ERD) in the first vaccine attempt and the difficulty in characterizing and stabilizing the pre-fusion F protein as a vaccine target.
View Article and Find Full Text PDFGMS J Med Educ
November 2024
Goethe University Frankfurt, Institute of General Practice, Frankfurt/Main, Germany.
Background: The attractiveness of general practice (GP) is increasing, as is evident in its growing popularity among students and the sharp rise in specialist certifications. However, in view of the future challenges at the GP level of care, there is a shortage of young doctors. It would thus be desirable if postgraduate training could be completed quickly, particularly in this area.
View Article and Find Full Text PDFSAGE Open Nurs
December 2024
Addis Ababa University College of Health Sciences, School of Nursing and Midwifery, Addis Ababa, Ethiopia.
Background: Continuous Professional Development (CPD) programs are crucial for enhancing the quality of healthcare delivery. The presence of incompetent and unskilled healthcare professionals is not satisfactory. This study aimed to evaluate nurses' engagement in CPD programs and identify the barriers affecting this engagement.
View Article and Find Full Text PDFEnter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!