作者: Markus Varsta , Jukka Heikkonen , Jouko Lampinen , José Del R. Millán
关键词:
摘要: This paper compares two Self-Organizing Map (SOM) based models for temporal sequence processing (TSP) both analytically and experimentally. These models, Temporal Kohonen (TKM) Recurrent (RSOM), incorporate leaky integrator memory to preserve the context of input signals. The learning convergence properties TKM RSOM are studied we show that is a significant improvement over TKM, because allows simple derivation consistent rule. results analysis demonstrated with experiments.