In this paper we describe a novel implementation of adaboost for prediction of survival function. We take different variations of the algorithm and compare the algorithms based on system run time and root mean square error. Our construction includes right censoring data and competing risk data too. We take different data set to illustrate the performance of the algorithms.