Bipolar Disorder (BD) has been linked to death and disability globally and is among the top ten causes of fragility in adolescents. Although BD is a common psychological disorder, it can occasionally be misinterpreted as depression, resulting in counseling those who are affected challenging. Machine Learning (ML) offers sophisticated abilities and methods for improved BD assessment. Examining ML techniques for recognizing and making a diagnosis of BD and its variants is the goal of this paper. Using Google Scholar, ScienceDirect, and PubMed as our three databases of choice. For the Recommended Objects to submit for Comprehensive Evaluations and Meta- Analysis (PRISMA), we succeeded to find 573 pertinent papers. The articles below were chosen according to how effectively their used machine learning (ML) and deep learning (DL). After evaluating every study independently, the recommended methods were grouped according to the various diseases that currently underway investigation was studying. Additionally, a list of numerous datasets that are freely accessible is provided, along with a discussion of the challenges the investigators faced. Irrespective of the individuals’ characteristics or whether or not they were juxtaposed with individuals with mental health diagnoses, this scoping overview offers a summary of recent research using ML systems for recognizing individuals who have BD.