Abstract
With the increasing prevalence of mental health issues, particularly suicidal behaviors, the need for early and accurate detection has become critical. This paper explores the current landscape of machine learning approaches used for the detection of suicidal tendencies. It examines a wide range of machine learning techniques applied to various data sources, including social media, clinical records, psychological assessments, self-reported forms like PHQ-9, audio speech recordings, and multimodal data integrating speech and visual information. This comprehensive review aims to reveal the types of existing research based on these varied datasets, highlighting the nuances of data collection, significant features identified, and the results obtained by different studies. Additionally, the review discusses the challenges and limitations associated with these approaches, providing researchers and practitioners with valuable insights into the potential and pitfalls of machine learning applications in diagnosing individuals at risk of suicide. The goal is to inform future research and improve early detection methods to ultimately reduce suicide rates.