Long term active learning from large continually changing data sets

摘要: Methods and systems are disclosed for autonomously building a predictive model of outcomes. A most-predictive set signals S k is identified out s 1, 2,..., D each one or more outcomes o . probabilistic models O = M ( ) learned, where prediction outcome derived from the that uses as inputs values obtained The step learning repeated incrementally data contains examples 1 , 2 ,..., corresponding K Various embodiments also apply to various physiological events autonomous robotic navigation.