Identification of Corrupted Data via $k$-Means Clustering for Function Approximation. CSIAM-AM 2021, 2 (1), 81-107. https://doi.org/10.4208/csiam-am.2020-0212.