Development of new diagnostic techniques: Machine learning

Research output: Chapter in Book/Report/Conference proceedingChapters

5 Citations (Scopus)


Traditional diagnoses on addiction reply on the patients’ self-reports, which are easy to be dampened by false memory or malingering. Machine learning (ML) is a data-driven procedure that learns algorithms from training data and makes predictions. It is quickly developed and is more and more utilized into clinical applications including diagnoses of addiction. This chapter reviewed the basic concepts and processes of ML. Some studies utilizing ML to classify addicts and non-addicts, separate different types of addiction, and evaluate the effects of treatment are also reviewed. Both advantages and shortcomings of ML in diagnoses of addiction are discussed. Copyright © 2017 Springer Nature Singapore Pte Ltd.
Original languageEnglish
Title of host publicationSubstance and non-substance addiction
EditorsXiaochu ZHANG, Jie SHI, Ran TAO
Place of PublicationSingapore
ISBN (Electronic)9789811055621
ISBN (Print)9789811055614
Publication statusPublished - 2017


Sun, D. (2017). Development of new diagnostic techniques: Machine learning. In X. Zhang, J. Shi, & R. Tao (Eds.), Substance and non-substance addiction (pp. 203-215). Singapore: Springer.


  • Addiction
  • Machines learning
  • Neuroimaging
  • Prediction
  • Training


Dive into the research topics of 'Development of new diagnostic techniques: Machine learning'. Together they form a unique fingerprint.