Investigating the understandability of XAI methods for enhanced user experience: When Bayesian network users became detectives

Butz, Raphaela; Schulz, Renée; Hommersom, Arjen; Eekelen, M.C.J.D. van

doi:10.1016/j.artmed.2022.102438

Cited by 5 publications

(4 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Starting from a training dataset for the target function and a new training instance ( X 1 ,X 2 , …,X m , m is the total number of instances), for which we must predict the C j class to which the instance belongs, through the Bayesian approach to the new instance we will assign it the most likely class. The most likely target C P can be expressed by (4) [ 41 ], [ 44 ]:

Where: P(C j ) is class probability and P(X i │C j ) is the probability of instance with number i , X i to be in class with number j, C j . So, we must discover the value of the C j class for which the value of the product is maximum for the new instance.…”

Section: Methodology For Identifying the Types Of Defects Of Bldc Motorsmentioning

confidence: 99%

Finding fault types of BLDC motors within UAVs using machine learning techniques

Andrioaia,

Gaitan

2024

Heliyon

View full text Add to dashboard Cite

Section: Methodology For Identifying the Types Of Defects Of Bldc Motorsmentioning

confidence: 99%

Finding fault types of BLDC motors within UAVs using machine learning techniques

Andrioaia,

Gaitan

2024

Heliyon

View full text Add to dashboard Cite

“…A study comparing some explanatory methods of BNs, even if they do not include counterfactuals, was conducted by Butz et al [13]. The purpose of the study is to evaluate the user experience of four different explanation approaches for Bayesian network inference in the medical domain.…”

Section: Related Workmentioning

confidence: 99%

“…Previous research [13] has shown that BNs themselves are perceived as an understandable representation of a case or situation. Participants felt that they were able to understand the explanation that was given directly from the BN.…”

Section: Introductionmentioning

confidence: 99%

Evaluating the Usefulness of Counterfactual Explanations from Bayesian Networks

Butz,

Hommersom,

Schulz

et al. 2024

Hum-Cent Intell Syst

Self Cite

View full text Add to dashboard Cite

Bayesian networks are commonly used for learning with uncertainty and incorporating expert knowledge. However, they are hard to interpret, especially when the network structure is complex. Methods used to explain Bayesian networks operate under certain assumptions about what constitutes the best explanation, without actually verifying these assumptions. One such common assumption is that a shorter length of the causal chain of one variable to another enhances its explanatory strength. Counterfactual explanations gained popularity in artificial intelligence over the last years. It is well-known that it is possible to generate counterfactuals from causal Bayesian networks, but there is no indication which of them are useful for explanatory purposes. In this paper, we examine how to apply findings from psychology to search for counterfactuals that are perceived as more useful explanations for the end user. For this purpose, we have conducted a questionnaire to test whether counterfactuals that change an actionable cause are considered more useful than counterfactuals that change a direct cause. The results of the questionnaire indicate that actionable counterfactuals are preferred regardless of being the direct cause or having a longer causal chain.

show abstract

“…A recent study compared an explainable-bydesign model (or "white box"), namely, a Bayesian Network (BN), with the explanations derived from XAI methods, by performing a survey with human participants. They found that BNs are easier to interpret compared to XAI methods [23]. Researchers have also emphasized that the belief in black box models surpassing explainable-by-design models in performance is unfounded and that XAI explanations may not consistently reflect the original model's computations.…”

Section: Introductionmentioning

confidence: 99%

Bayesian Networks in the Management of Hospital Admissions: A Comparison between Explainable AI and Black Box AI during the Pandemic

Nicora,

Catalano,

Bortolotto

et al. 2024

J. Imaging

View full text Add to dashboard Cite

Artificial Intelligence (AI) and Machine Learning (ML) approaches that could learn from large data sources have been identified as useful tools to support clinicians in their decisional process; AI and ML implementations have had a rapid acceleration during the recent COVID-19 pandemic. However, many ML classifiers are “black box” to the final user, since their underlying reasoning process is often obscure. Additionally, the performance of such models suffers from poor generalization ability in the presence of dataset shifts. Here, we present a comparison between an explainable-by-design (“white box”) model (Bayesian Network (BN)) versus a black box model (Random Forest), both studied with the aim of supporting clinicians of Policlinico San Matteo University Hospital in Pavia (Italy) during the triage of COVID-19 patients. Our aim is to evaluate whether the BN predictive performances are comparable with those of a widely used but less explainable ML model such as Random Forest and to test the generalization ability of the ML models across different waves of the pandemic.

show abstract

Investigating the understandability of XAI methods for enhanced user experience: When Bayesian network users became detectives

Cited by 5 publications

References 15 publications

Finding fault types of BLDC motors within UAVs using machine learning techniques

Finding fault types of BLDC motors within UAVs using machine learning techniques

Evaluating the Usefulness of Counterfactual Explanations from Bayesian Networks

Bayesian Networks in the Management of Hospital Admissions: A Comparison between Explainable AI and Black Box AI during the Pandemic

Contact Info

Product

Resources

About