ABSTRACT
The widespread use of wearable electronic devices for health monitoring is only a matter of time. Biomedical algorithms are enabling factors for the solutions, as they are required for the interpretation of the physiological signals acquired from the users. This chapter provides an overview in terms of the problem types and how each can be practically formulated. The required outcomes are discrete and the parameters of the classification model are fixed after the training stage, and cannot be altered during run-time. The exploration of new or existing approaches involves a comprehensive review to identify existing algorithms, which can be used to solve the formulated problem in the light of existing constraints. The most promising candidates are then ported over to a programming language compatible with the target device. The dataset used in the training and evaluation of algorithm is crucial, as it impacts the algorithmic parameters, the relative results between algorithm candidates and potentially the choice of algorithm.
