Skip to main content
EURASIP Journal on Audio, Speech, and Music Processing
Figure 2 | EURASIP Journal on Audio, Speech, and Music Processing

Figure 2

From: Decision tree-based acoustic models for speech recognition

Figure 2

A split based on a question that maximizes the increase in log likelihood. N T and N all are the number of true samples reaching node d and the total number of samples (true and false) assigned to node d. N T y and N a l l y are the numbers of the true and all samples at child yes, N T n and N a l l n are the numbers of the true and all samples at child no, respectively

Back to article page