A proposal for variable selection in the Cox model

We propose a new method for variable selection and estimation in Cox's proportional hazards model. Our proposal minimizes the log partial likelihood subject to the sum of the absolute values of the parameters being bounded by a constant. Because of the nature of this constraint it tends to produce some coeecients that are exactly zero and hence gives interpretable models. The method is a variation of the \lasso" proposal of Tibshirani (1994), designed for the linear regression context. Simulations indicate that the lasso can be more accurate than stepwise selection in this setting.