Brain signal data are records of the electrical activity of the brain made with electroencephalography (EEG). EEG data mining is one of the key topics which has received lots of attention. State-of-art machine learning (ML) methods are incompetent to dive deep and learn the EEG data pattern. Moreover, additional feature extraction methods are employed for feature extraction when a classical ML is in action which adds up additional computational burden. Lastly, the multi-class EEG data classification performance is not quite satisfactory. To overcome these issues, we propose a computerized framework for efficient and effective irregular brain signal data recognition network (IBDR-Net). Our suggested IBDR-Net is made up of four steps: (1) EEG data acquisition, (2) Pre-processing, (3) Recognition of irregular brain signal data using IBDR-Net, and (4) Performance evaluation. IBDR-Net framework has been tested and trained with real human EEG dataset and this study has set a new milestone for the EEG researchers and boosted the multi-class performance.