neural network

From Aaushi
Jump to navigation Jump to search

Introduction

  • group of interconnected units (neurons) that send signals to one another
  • units can be either biological cells or mathematical models[1]
  • in machine learning, a neural network is an artificial mathematical model used to approximate nonlinear functions
  • neural networks are usually arranged into layers, with information passing from the input layer through one or more intermediate hidden layers to the final output layer
  • the input to each unit is a number representing a linear combination of the outputs of the connected neurons in the previous layer
  • the behavior of the network depends on the strengths of connections between units
  • a network is trained by modifying these strengths of connections through empirical backpropagation in order to fit some preexisting dataset[1]
  • one of the problems of neural networks is over training
    • one might think the more the training, the better the fit to the dataset
    • not so, comes a point where neural networks trained further deteriorate in performance

More general terms

References