In our research, we focus on the task of classification for silent speech recognition with the goal of developing a brain-computer interface (BCI) based on electroencephalographic (EEG)
data, which will be capable of assisting people with mental and physical disabilities and expanding human capabilities in everyday life. Our previous research has shown that silent pronouncing some words results in almost identical distributions of electroencephalographic signal data. Such phenomenon has a suppressive impact on the quality of neural network models behavior. In this paper, we propose a data processing technique which distinguishes between statistically remote and inseparable classes in the dataset. Application of the proposed approach helps us to reachг the goal of maximizing the semantic load of the dictionary used in BCI.
Furthermore, we propose the existence of a statistical predictive criterion for the accuracy of binary classification of the words in a dictionary. The aim of such criterion is to estimate the lower and the upper bounds of classifiers’ behavior only by measuring quantitative statistical properties of the data (in particular, using the Kolmogorov-Smirnov method). We show that higher levels of classification accuracy can be achieved by means of applying the proposed predictive criterion,
making it possible to form an optimized dictionary in terms of semantic load for the EEG-based BCIs. Using such dictionary as a train dataset for classification tasks grants the statistical
remoteness of the classes by taking into account the semantic and phonetic properties of the corresponding words and improves the classification behavior of silent speech recognition models.