Developing machine learning algorithms for improved diagnosis and prognosis

Machine learning research has advanced quickly within the last decade, utilizing the availability of large data sets and data storage and processing advancements to develop state-of-the-art algorithms that rival and often outperform traditional statistical methods. Yet, while ML has been successfully applied in many fields of research from varying disciplines, including neuroscience, politics, criminology, ecology, and remote sensing, among many others, few biomedical researchers have explored the use of ML for improved disease diagnosis and prognosis. This is surprising, considering the fascinating and complex nature of disease, as well as the generalizability and flexibility of ML algorithms. In this chapter, we elucidate some key ML concepts in an effort to empower and excite the biomedical researcher community to further investigate the potential of utilizing ML techniques in diagnostics and prognostics.