Review on Artificial Intelligence Techniques for Medical Diagnosis

In this paper, different techniques used for medical diagnosis are discussed and their performances are compared. These techniques are CNN, Random forest, KNN, SVM etc. CNN (Convolution neural Network) includes layers and weighted mechanism to decide features that leads to correct classification. Random forest generates decision trees based on selected features. Highest frequency patterns are placed at root node and consequences are placed at child node using this mechanism. SVM is most frequently used mechanism in the field of disease prediction. It has two hyper planes; one indicates disease and another indicates non-disease features. KNN is based on Euclidean distance. Each of this mechanism is demonstrated experimentally and CNN appears to best in terms of classification accuracy.