1 paper · 1 filter
Amit Deshpande, Rameshwar Pratap
The k-means is a popular clustering objective, although it is inherently non-robust and sensitive to outliers. Its popular seeding or initialization called k-means++ uses $D^{2…