Neural Network in Asaya
By analysing the performance of millions of different feature-combinations and feature scaling, the neural network will not only extract relevant features, but will also engineer new relevant features based on the existing data.
The use of boosting algorithms and neural network will enable Asaya platform to indicate, which of the patient-features have a higher correlation with treatment outcome and use then accordingly. For example, given a patient’s medical history, Asaya platform will tell that a specific disease (i.e. type-II diabetes) is more associated with treatment outcome than another (i.e. psoriasis).