CVLoglikelihood

This function calculates the cross validated log likelihood score of a validation set. The score is calculated in the context of the training data by means of subtraction. See Houwelingen et al., 2006 for more information.

Parameters:

• model - the model returned by trainCoxMlp
• x_full - `2D numpy array of the training and validation set
• ytime_full - time of death or censoring for each patient in both training and validation set
• ystatus_full - censoring of each patient in both training and validation set
• x_train - 2D numpy array of the training set
• ytime_train - time of death or censoring for each patient in the training set
• ystatus_train - censoring of each patient in the training set

Reference: van Houwelingen, Hans C., et al. "Cross-validated Cox regression on microarray gene expression data." Statistics in medicine 25.18 (2006): 3201-3216.