[Retracted] The Mongolian Vowel Acoustic Model Based on the Clustering Algorithm

Table 4

Composition of training sentences and test sentences in invisible three-tone subexperiment.

Subjects	Number of training sentences	Training the sex ratio	Number of test sentences	Test number of words
Mongol 1	105  30 = 3150	1.5	20  30 = 600	390  30 = 11700
Mongol 2	100  30 = 3000	1.5	18  30 = 540	240  30 = 7200
Mongol-all	6150	1.5	1140	18900
Dialogue 1	325  45 = 14625	1.5	65  45 = 2925	306  45 = 13770
Dialogue 2	318  37 = 11766	1.47	63  37 = 2331	313  37 = 11581
Dialogue 3	320  30 = 9600	1.5	64  30 = 1920	371  30 = 11130
Dialogue-all	35991	1.49	7176	36481