SOM (self-organizing map) is modified to encode the cross-bifrequency map that evaluates the cardiorespiratory interactions at Zen-meditation and resting states. The effectiveness of meditation on stress manipulation has aroused the attention of researchers. Cardiovascular and respiratory systems are very important in manipulating both mental and physical health in the stressful environment. Cross-bispectral analysis, applied to heart-rate (HR) and respiratory (RP) sequences, quantifies the degree of phase coupling between different spectral components of two signals. The resulted cross-bifrequency map reveals the property of quadratic coupling between two oscillators, the cardiac and respiratory systems. Conventional SOM weights equally each attribute of the input (feature) vector. However, some attributes of the input vector may play a more decisive role in recognition task. We accordingly propose the input-weighted SOM (IwSOM). According to the IwSOM classification, rest session of control group is characterized by a series of coupling frequencies of which the respiratory rate (RR) is in resonance with HR at 0.05 Hz. The Zen-meditation bifrequency exhibits strong resonance between HR and RP in the same low frequency range.