Solving Economic Problems Applying Artificial Neural Networks

Abstract

Deep learning extends machine learning by using artificial neural networks to operate a set of predictor variables and predict future response variables. Artificial neural networks are a group of nodes that receive input values in the input layer and transform them in the subsequent hidden layer (a layer in between the input and output layer). This hidden layer transforms the nodes and allots varying weights (vector parameters that determine the extent of influence input values have on output values) and bias (a balance value which invariably is 1). Following that, they generate a set of output values in the output layer by applying an activation function. Artificial neural networks are part of deep learning, which advances machine learning through structuring models and by replicating human neural activity as the base. Propagation is the process of training networks (often through backward propagation, which involves updating weights in reserves, from the output layer).