Robust and Sparse Regression via $\gamma$-divergence

Takayuki Kawashima, Hironori Fujisawa

In high-dimensional data, many sparse regression methods have been proposed. However, they may not be robust against outliers. Recently, the use of density power weight has been st…