A Network of such nodes creates a Neural Network.

DEEP NEURAL NETWORK: When multiple hidden layers are used the Network is called Deep Neural Network.

THE EXPLOSION: Neural Network has achieved tremendous success in the area of Supervised Machine Learning. Just the introduction of the non-linear link function created high accuracy machine learning models running on big data. Deep Learning took it further ahead introducing more and more hidden layers.

THE INEVITABILITY: With the advent of cheap as well as high capacity data storage and retrieval systems orthodox statistical data analysis methods were failing to perform well. As per my understanding non-statistical techniques like neural network, deep learning were inevitable.

QUESTION I: Does Neural Network accept a Linear Link Function?
QUESTION II: Could Neural Network be used in Time Series Forecasting?