Skip to main content
. Author manuscript; available in PMC: 2008 Feb 5.
Published in final edited form as: J Neurophysiol. 2006 Oct 4;97(3):2525–2532. doi: 10.1152/jn.00190.2006

TABLE 1.

Most common trained HMM from raw detector output training sequences


A=[0.95940.04068.08×1060.24890.75110.00010.00100.00350.9955]
B=[0.95400.04600.00.06270.93730.00.00.01.0]
S=[0.85670.13970.0036]

This HMM was found 78% of the time (39 of 50 random starts) with the constraints outlined in methods. This HMM was also the one associated with the highest probability of producing the training sequences. A, transition probability matrix; B, symbol emission probability matrix; S, 3 × 1 vector of s1, s2, and s3.