Artificial intelligence and machine learning

Abstract

The terms artificial intelligence and machine learning are often used interchangeably but have distinct meanings. Artificial intelligence is the broader term and refers to a range of innovations from domains such as robotics, computer vision, speech and natural language processing, and machine learning. One weakness of machine learning models is their propensity for over-fitting to the training data. Highly structured data present the least challenge to machine learning research. In each case, the data has a highly standardised inherent structure that makes it easier to encode for a machine learning model. A standard data model also facilitates de-identification of data when clinical data must be transferred to an academic or commercial partner for development of the machine learning model. From the perspective of a machine learning researcher and a healthcare innovator, it is important to note a key transition that occurs in these four problem types.