Table 5.

List of studies that used local interpretable model agnostic explanations.

Authors	Year	Objective	Data set(s)	Data type	Important features	Machine learning (ML)/Deep learning (DL)	Explained model
Kaplun et al. [90]	2021	Extract complex features from cancer cell images and classify malignant and benign cancer cell images	BreakHis [91]	Microscopic images	Yellow highlighted segments in the image	DL	ANN (2‐layer feed forward neural network)
Saarela et al. [92]	2021	Comparing different feature importance measurements using linear (LR) and nonlinear (RF) classification ML models	Breast Cancer Wisconsin (Diagnostic) [87]	Text	L1‐LR $\to$ all except one (compactness 3) RF $\to$ nine features were significant	ML	L1 regularized LR, RF
Adnan et al. [93]	2022	Proposing a model in BC metastasis prediction that can provide personalized interpretations using a very small number of biologically interpretable features	Amsterdam Classification Evaluation Suite (ACES) [94] (composed of 1616 patients, among which 455 is metastatic)	Genomic data	N/A	M/DL	RF, LR, lSVM, rSVM, ANN
Maouche et al. [95]	2023	Propose an explainable approach for predicting BC distant metastasis that quantifies the impact of patient and treatment characteristics	Public data set composed of 716 Moroccan women diagnosed with breast cancer [96]	Clinicopathological data	The characteristics have different impacts ranging from high, moderate, and low	ML	Cost‐sensitive CatBoost
Deshmukh et al. [97]	2023	Improve the qk‐means clustering algorithm using LIME to explain the predictions	The breast cancer data set has 600 attributes or patient records and 7 features	Text	A tabular explainer explains the positively and negatively correlated features	ML	qk‐means (hybrid classical‐quantum clustering approach)