© 2022 IEEE.Multi-view Brain Networks (MBNs) have made a significant contribution to the diagnosis of neurological disorders using computers. MBNs can boost the classification accuracy of neurological disorders such as Alzheimer's disease (AD) by capturing diverse characteristics from multiple perspectives of a population. MBNs provide rich data diversity and better generalization for the classifier models, but the classifiers require large-sized dataset to avoid being over-trained. Moreover, it is challenging to collect clinical data because it is scarce and expensive. In this work, we present a method that augments fake (i.e., synthetic) data samples to boost the neurological disorder classification models. It has been shown that the classifiers trained on our synthetic samples achieve better results than those trained on synthetic samples of traditional data augmentation techniques.