Supervised Learning Algorithm of Classification on Basis of Ranges

Abstract

A supervised learning algorithm of classification which is implemented on linear data. Ranges/Boundaries of each group are calculated. As the data is linear the number of ranges will be equal to a number of classes. New input data will be tested by in which range/boundary it lies. And the range in which it lies, the class of that specific range will be assigned to the new data. The experimental results show that the accuracy depends on the linearity of the data. Therefore our model can show accuracy up to 99.99% also if the data is completely linear in nature.