JMIR Publications

ABSTRACT

The concept of applying machine learning to perform analysis for unearthing meaningful features, the success mainly depends on the ability to accomplish the intended task. Currently, devices and applications are capable of generating hundreds of millions of data in the form of images, audio, text, graph, speech and signals creating the concept of Big Data. The innovation of Deep Learning (DL) is a developing trend in the wake of big data for data representation and analysis. DL is a type of machine learning that has deeper (or more) inner hidden layers of similar function cascaded into the network and has the capability to make meaning from big data. Technological advancement, software implementation and successive impressive result which have been reported has contributed to wide acceptance of DL. The increasing and widespread application of deep learning in different discipline is overwhelming and the success story is impressive. This review paper highlights the fundamentals of DL models and presents a general view of the trends in deep learning by capturing literature from PubMed and IEEE database publication that implements different variants of DL models. This paper also presents some inherent challenges in deep learning, prospective research directions that focuses on improving health management by promoting the application of physiological signals and modern internet technology.