The purpose of the stay was to deepen in some of the topics of the thesis, whose framework is nonparametric estimation of the probability of default in credit risk. The stay took place at the Katholieke Universiteit Leuven (KU Leuven) under the supervision of Professor Ingrid Van Keilegom.The research carried out during the stay was based on the use of cure models to estimate the probability of default. Cure models have been used in different contexts, but mainly medical. More recently, cure models have been introduced to model problems in credit risk, given that a large proportion of the population may never experience the event of interest, the default, in the duration of their credits. Several researchers, including Professor Ingrid Van Keilegom, have proposed conditional survival function estimators based on mixed cure models. During this stay, such estimators were used to estimate the probability of default. It is important to take into account that the default probability function can be written as a transformation of the survival function. Desirable asymptotic properties of the resulting probability of default estimators were proved and their behaviour was analysed by simulation. In addition, the simulations carried out allowed us to compare the behaviour of the probability of default estimators obtained by means of cure models with the nonparametric estimators considered up to that moment in the thesis. This study concludes with a brief analysis of credit data illustrating the use of these estimators. The probability of default is estimated for a set of credits granted by a German financial institution.