In large-scale Tokamaks disruptions have the potential to create serious damage to the facility. Hence disruptions must be avoided, but, when a disruption is unavoidable minimizing its severity is mandatory. A reliable detection of a disruptive event is required to trigger proper mitigation actions. To this purpose machine learning methods have been widely studied to design disruption prediction systems on several experimental devices. In particular, for ASDEX Upgrade, some of the authors presented predictive systems applying data based techniques, such as Multi-layer Perceptron neural network [1], Discriminant Analysis [2], and Self-Organizing Maps [3]. The training phase of the proposed approaches is based on the availability of disrupted and non-disrupted discharges. To accomplish an exhaustive model every disruptive and safe configurations included in the machine operational space should be represented in the training set. Safe configurations were selected from safe discharges, while disruptive configurations were assumed appearing into the last 45ms of each disruption [1,3]. Even if the achieved results in terms of correct predictions were good, it has to be highlighted that the choice of such a fixed temporal window might have limited the prediction performance. In fact, it generates ambiguous information in cases of disruptions with disruptive phase shorter than 45ms. Conversely, missing information is caused in case of disruptions with a disruptive phase longer than the prefixed one. The assessment of a specific disruptive phase for each disruptive discharge represents one of the most relevant issues in understanding the disruptive events. Several similarity measures, such as Mahalanobis distance, and statistical methods, such as Logistic Regression, have been applied to evaluate the membership of each sample to the safe or the disruptive configurations. Preliminary results show that enhancements on the achieved performance on disruption prediction are possible by defining a specific disruptive phase for each disruption.