Article directory
Neural Networks
-
Neural network: a network structure of a large number of neuron nodes connected according to a certain architecture - brain structure
-
The role of neural networks
- classification
- pattern recognition
- continuous value prediction- Establish a mapping relationship between input and output
biological neurons
artificial neuron
Each neuron is an independent unit with a similar structure, accepting the data from the previous layer, and inputting the weighted sum of these data into the nonlinear action function, and finally passing the output of the nonlinear action function to the next layer .
activation function
Derivative
Artificial neural networks
Common understanding of "layer"
feedforward neural network
A type of artificial neural network that has no feedback and can be represented by a directed acyclic graph.
Delta learning rule
A supervised learning algorithm. The connection weight is adjusted according to the difference between the actual output of the neuron and the expected output.
Objective function of feed-forward neural network
gradient descent
Output layer weight change amount
Error Direction Propagation Algorithm
Error Propagation Iterative Formula
Simple BP calculation example
Stochastic Gradient Descent (SGD)
Mini-batch Gradient Descent
Typical machine learning steps
The impact of features on learning
Features of Deep Learning
Deep learning is an extension of the neural network model.
The Rise and Fall of "Connectionism"
The Difference Between Deep Learning and Neural Networks
