The effect of multi-instance learning on hybrid classification performance of EEG and NIRS data

Brain-Computer Interface (BCI) is a system that enables the signals obtained as a result of neural activities in the person’s brain to be processed as commands using a computer system. BCI system consists of a user, computer, and peripherals. EEG and NIRS are the primary imaging systems for representative brain signals. The performance of BCI systems is directly proportional to the classifier’s performance. Since BCI is an emerging technology, especially hybrid studies are limited. Hybrid systems are used to overcome the limitations of one-sided systems and increase the accuracy of the classifier. This study considered the binary classification of Word Generation (WG) and Baseline (BL) cognitive tasks from BCI tasks. After feature extractions were made on an open dataset, Multi-Instance Learning (MIL) was applied, and the performances of various classifiers were measured. Feature extraction operations were done in the time domain and frequency domain. Fusion operation performed at the feature level affected the performance positively. Classification results at the level of 99% showed that the MIL method would lead to future studies.