Machine learning in autism spectrum disorder diagnosis and treatment: techniques and applications

Machine learning (ML) has become increasingly useful in health care, demonstrating effectiveness in a wide range of tasks such as diagnosing conditions and recommending treatment. ML has also been applied to study autism spectrum disorder (ASD), and it has the potential to transform the methods by which ASD is diagnosed and treated. This chapter provides a brief review of ML applications in ASD research. Three main fields of ML-ASD literature are covered: (1) use of ML in diagnosing ASD, (2) use of ML in analyzing characteristics of ASD, and (3) use of ML in real-world applications to diagnose and treat ASD. Several studies, their approaches, their success, as well as their drawbacks are discussed in each section. We also provide recommendations for future research based on trends identified in the literature. We hope this review provides a useful background for ML research to continue making significant strides relating to ASD.