Introduction to Keras
Keras is a Google platform and a high-level interface to build ML/DL models with TensorFlow. Keras provides a high-level API that encapsulates data transformations and operations using logic units, called layers, as the building blocks to create neural networks. A layer performs data manipulation operations such as taking an average, calculating the minimum, and so on.
With Keras, ML models are built from layers. During the ML model training process, the variables in the layers are adjusted, via backpropagation, to optimize the model cost function. Behind the scenes, TensorFlow and Keras complete detailed data operations, such as linear algebra and calculus calculations, in the background. Keras provides the following two APIs:
- The sequential API provides the simplest interface and least complexity. With the sequential API, we can create the model layer by layer and thus build an ML/DL model as a simple list of layers.
- The functional API is more...
