KERAS is the high level of API which is embedded in the Tensor Flow. It is being used to be user friendly, modular extendable and this can help us while working in Python. It was initially designed to ease the working process of human being and not for machine.
Our aim is to reduce the cognitive load with the application of KERAS. The chief reason to use KERAS originate from its guiding principles of user friendly apart from its ease of learning. It consists of synthesis of five engines such as Tensor Flow, CNTK, Theano, MX Net and Plaid ML. Moreover, it is supported by Google, Microsoft, Amazon, Apple, Uber and others.
KERAS would supply seven common deep learning data sets including Cifar 10 and Cifar 100 for colour images, PMB movie reviews, MNIST images etc.