Neural Network

A neural network is a Probabilistic Algorithm created by Differentiable Programming comprised of nodes with weights and biases generated.

Types

• Recurrent Neural Networks
• Convoluted Neural Network

Theory

A neural network consists of Layers of Neurons that are activated depending on their Activation Function and the total input value calculated by the Weights,Bias of previous input layers. This process is repeated until we compute the output neurons in the final layer for practical use. Weights and Biases are later refined during Gradient Descent by Back Propagation in the training phase.

Concepts

• Neuron
• Layer
• Activation Function
• Biases
• Weights
• Loss Function
• Gradient Descent
• Back Propagation
• Stochastic Gradient Descent