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