Fig. 1.

Basics of deep learning models. (a) Deep learning is a subset of the machine learning approach, which itself is a subdomain of artificial intelligence. The amount of dots in boundary line schematically represents amount of datasets available for model training (b) A basic unit of artificial neural network is a neuron, which can take input and undergo calculations involving weight and biases. The resulting values are subjected to an activation function to decide whether or not the neuron's input to the network is significant in the prediction process utilizing simpler mathematical procedures. (c) Deep learning models consist of multiple layers of such indivisual neurons with the ability to process high-dimension data. The dataset is fed into the network in the input layer, which is subsequently followed by multiple inner hidden layers, and final outcomes are computed on the outer layer with the utility of different activation functions.