A Machine Learning Assessment System for Spoken English Based on Linear Predictive Coding

Figure 10

Percentage of 3 types of pronunciation errors.