Explainable Predictive Process Monitoring

Galanti, Riccardo; Coma-Puig, Bernat; Leoni, Massimiliano de; Carmona, Josep; Navarin, Nicolò

doi:10.48550/arxiv.2008.01807

Cited by 1 publication

(3 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…State-of-the-art explainable predictive process analytics have generally attempted to use existing explainable methods in XAI. For example, the use of LIME and SHAP to evaluate and improve black box models [13,14] has been explored, and SHAP has been used to create explainable dashboards for informed decisionmaking [3]. A method of interpreting process predictive models underpinned by neural networks using layer-wise relevance propagation is presented in [19].…”

Section: Explainable Predictive Process Analyticsmentioning

confidence: 99%

“…SHAP and LIME, two popular post-hoc interpretation methods, are chosen as the explainable methods to be evaluated, as they have previously been used in predictive process analytics literature [3,13,14].…”

Section: Design Of Experimentsmentioning

confidence: 99%

“…Several recent studies in predictive process analytics have attempted to apply existing explainable methods in XAI to interpret black box models used to compute process predictions. For example, LIME and SHAP have previously been used to interpret process predictive models built on complex machine learning algorithms [3,13,14]. While it is important to understand how well the explanations generated by an explainable method may be received by end users, it is essential to study how fit for purpose such an explainable method is when being used to explain predictive models for process predictions, and how the characteristics of datasets and underlying machine learning techniques affect the quality of explanations generated.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Evaluating Explainable Methods for Predictive Process Analytics: A Functionally-Grounded Approach

Velmurugan¹,

Ouyang²,

Moreira³

et al. 2020

Preprint

View full text Add to dashboard Cite

Predictive process analytics focuses on predicting the future states of running instances of a business process. While advanced machine learning techniques have been used to increase accuracy of predictions, the resulting predictive models lack transparency. Current explainable machine learning methods, such as LIME and SHAP, can be used to interpret black box models. However, it is unclear how fit for purpose these methods are in explaining process predictive models. In this paper, we draw on evaluation measures used in the field of explainable AI and propose functionally-grounded evaluation metrics for assessing explainable methods in predictive process analytics. We apply the proposed metrics to evaluate the performance of LIME and SHAP in interpreting process predictive models built on XGBoost, which has been shown to be relatively accurate in process predictions. We conduct the evaluation using three open source, real-world event logs and analyse the evaluation results to derive insights. The research contributes to understanding the trustworthiness of explainable methods for predictive process analytics as a fundamental and key step towards human user-oriented evaluation.

show abstract