An overview of the activation functions used in deep learning algorithms

In deep learning models, the inputs to network are processed using activation functions generate output corresponding these inputs. Deep models of particular importance in analyzing big data with numerous parameters and forecasting useful for image processing, natural language object recognition, financial forecasting. Also, algorithms, have been developed by taking into account features such as performing process a healthy way, preventing excessive learning, increasing accuracy performance, reducing computational cost. this study, we present an overview common current used algorithms. fixed trainable introduced. As functions, sigmoid, hyperbolic tangent, ReLU, softplus swish, LReLU, ELU, SELU RSigELU