2012 | OriginalPaper | Buchkapitel
Neural Network Analysis of Right-Censored Observations for Occurrence Time Prediction
verfasst von : Young U. Ryu, Jae Kyeong Kim, Kwang Hyuk Im, Hankuk Hong
Erschienen in: E-Life: Web-Enabled Convergence of Commerce, Work, and Social Life
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Introduced is a neural network method to build survival time prediction models with censored and completed observations. The proposed method modifies the standard back-propagation neural network process so that the censored data can be used without alteration. On the other hand, existing neural network methods require alteration of censored data and suffer from the problem of scalability on the prediction output domain. Further, the modification of the censored observations distorts the data so that the final prediction outcomes may not be accurate. Preliminary validations show that the proposed neural network method is a viable method.